WEBVTT
00:00:05.120 --> 00:00:08.000
Hello everyone and welcome to the Source Forge podcast.
00:00:08.080 --> 00:00:09.679
I'm your host, Bo Hamilton.
00:00:09.839 --> 00:00:17.679
Now, for a long time, hiring has leaned heavily on resumes, interviews, and a little bit of, I would say, gut instinct.
00:00:17.760 --> 00:00:21.600
Uh, but in the AI era, that whole equation is starting to change.
00:00:21.679 --> 00:00:27.039
You know, there are a lot of ways candidates are are using AI to refine their application materials these days.
00:00:27.120 --> 00:00:37.200
They can tailor their experience to a job description or even prepare super polished answers in advance, both of which are, you know, understandable from the candidate's perspective.
00:00:37.359 --> 00:00:41.359
You want to stand out from the competition and come across as well as possible.
00:00:41.439 --> 00:00:47.520
Um, but from the hiring perspective, the resume, you know, starts to become a little bit less of a signal.
00:00:47.600 --> 00:00:49.920
Um, it's a weaker signal than it used to be.
00:00:50.000 --> 00:00:57.439
Um, hiring managers have to kind of cut through the AI fluff, so to speak, to find the right candidate with the right know-how for the job.
00:00:57.600 --> 00:01:02.079
And so that is why more companies are shifting towards what's called skills-based hiring.
00:01:02.159 --> 00:01:07.760
Um, and they're asking a more direct question of just can this person actually do the work that they say they can do?
00:01:07.920 --> 00:01:10.239
Um, and so that's where you know our guest of the show comes in.
00:01:10.319 --> 00:01:14.239
Today I'm joined by Nick Leslie, director of customer success at Candy Tech.
00:01:14.480 --> 00:01:29.200
Uh, it's an AI-powered skills assessment platform built around job simulations that help companies assess candidates across a variety of metrics, you know, hard skills, soft skills, leadership, uh tests for behavioral and cognitive traits, and even, of course, AI skills as well.
00:01:29.359 --> 00:01:32.079
That's an increasingly important skill set in today's world.
00:01:32.159 --> 00:01:33.519
So we've got a lot to talk about.
00:01:33.599 --> 00:01:41.760
Uh, we're gonna talk about how AI is impacting the hiring process, what good job simulations actually look like, and then what strong candidates look like in the first place.
00:01:41.920 --> 00:01:44.319
So, again, we've got a lot to cover.
00:01:44.480 --> 00:01:46.560
So, without further ado, let's get right into it.
00:01:46.640 --> 00:01:48.079
Nick, welcome to the podcast.
00:01:48.239 --> 00:01:49.040
Glad you could join us.
00:01:49.200 --> 00:01:49.439
Thank you.
00:01:49.599 --> 00:01:50.640
Appreciate you having me on.
00:01:50.879 --> 00:01:51.519
Absolutely.
00:01:51.680 --> 00:02:02.000
Um, so I actually had the pleasure of interviewing your CEO, Guy Burrell, last year uh to get an understanding of you know the solutions your company is working on to help tackle the, you know, in the hiring space.
00:02:02.159 --> 00:02:09.280
Um, for listeners who might have missed that episode, help us understand the company and kind of the lens you're coming from.
00:02:09.680 --> 00:02:10.159
Yeah, for sure.
00:02:10.319 --> 00:02:12.080
Well, so about it again, thank you for having me on.
00:02:12.159 --> 00:02:14.719
Um, I'm Nick, director of Custom Success at Candy Tech.
00:02:14.960 --> 00:02:23.680
Um, and effectively, what Candy Tech does, we're an all-in-one skills assessment platform that allows you to see how someone would perform a job before they get a job.
00:02:23.919 --> 00:02:31.120
Really, what we believe in is creating simulations that mimic tasks that candidates would have to perform if they were employed in your role.
00:02:31.199 --> 00:02:36.479
So you can get a sense of can they do what you need them to be able to do within your environment?
00:02:36.639 --> 00:02:38.879
And we're extremely holistic in the way that we work.
00:02:38.960 --> 00:02:54.080
So we don't just cover one function of a business and we can run skill-based assessments across any area that you're hiring, whether that's finance, data, marketing, legal, of course, technology, um, which is becoming very essential for people to assess somebody's skills.
00:02:54.240 --> 00:02:58.080
And you can translate assessments into any language on the planet as well.
00:02:58.159 --> 00:03:10.639
So if you're hiring uh really in any geography, you're able to run assessments to get a strong signal if a candidate is worth you investing time into to get into the process and ultimately to get a quality higher.
00:03:10.960 --> 00:03:11.360
Got it.
00:03:11.520 --> 00:03:13.599
Okay, thanks for providing that that overview.
00:03:13.759 --> 00:03:22.000
I think it's uh, you know, what you're doing is more important than ever, you know, because I think people people have been talking about how resumes are just becoming less useful um for a while.
00:03:22.159 --> 00:03:27.439
Now that's been kind of the the conversation I've been hearing amongst friends and family who are looking for jobs.
00:03:27.599 --> 00:03:35.280
Um I know that a resume, let's say I think it's fair to say it was never a perfect measure of whether someone could actually do the job.
00:03:35.360 --> 00:03:55.840
Um, but now with AI, it's e it's even easier for candidates to, you know, do things like we talked about, a polishing resume, uh polishing their experience, using specific language that is tailored to the job description and uses really specific keywords and just really just makes them appear stronger than they actually are on paper, um, than they might be in practice.
00:03:56.000 --> 00:04:01.680
So um, yeah, you can you can check a lot of boxes, look great on paper, but you don't really necessarily have the right candidates.
00:04:01.840 --> 00:04:10.000
So I'm curious, in this AI world, how has your thinking changed on on what you know world uh world beyond resumes actually looks like?
00:04:10.319 --> 00:04:13.120
Yeah, and actually you've you've taken a lot of uh my thoughts.
00:04:13.199 --> 00:04:15.120
Uh you said a lot of my thoughts out loud just Sarah.
00:04:15.199 --> 00:04:17.279
Actually, it's really interesting to hear how you describe that.
00:04:17.360 --> 00:04:19.439
I've been working in hiring now for 16 years.
00:04:19.519 --> 00:04:22.160
So I've seen things shift and evolve a lot.
00:04:22.319 --> 00:04:25.839
And I've always thought, as you just said, the resumes are somewhat imperfect.
00:04:26.000 --> 00:04:34.399
You know, they're self-reported, they rely heavily on someone's past achievements, and fundamentally they just don't tell you how someone will perform in a role.
00:04:34.560 --> 00:04:45.759
But now, actually, with the advancements of AI, my thinking shifted from them being imperfect to being increasingly unreliable, particularly as a primary signal of job performance.
00:04:45.839 --> 00:04:57.360
And this has created a very unique challenge for TA because if we look at the broader environment, as you've just said, right now candidates can tailor their CV to pretty much any role they're applying to instantly.
00:04:57.439 --> 00:04:59.680
And they can do it at scale for hundreds of applications.
00:04:59.839 --> 00:05:03.040
So companies are seeing two things happening at the same time.
00:05:03.199 --> 00:05:06.079
One is that application volume is going through the roof.
00:05:06.240 --> 00:05:10.000
And two, in parallel, a lot of resumes are starting to look very similar.
00:05:10.160 --> 00:05:16.240
I hear constantly from our customers that every candidate or every resume they see basically looks the same.
00:05:16.319 --> 00:05:20.160
So it's really hard to identify who should we focus on and who should we get into process.
00:05:20.319 --> 00:05:22.399
And they're getting hundreds of them for every role.
00:05:22.560 --> 00:05:32.480
So I think that's why the conversation is shifting now from how do we screen resumes better to should resumes even be the thing that we rely on in the first place?
00:05:32.639 --> 00:05:36.319
And that's where skills-based hiring is really coming into the conversation.
00:05:36.480 --> 00:05:42.480
The way we do this at Candy Tech, uh and what we see working really well, is giving candidates something that really mirrors the job.
00:05:42.639 --> 00:05:48.959
So a simulation of the role and the company's environment and then seeing how the candidate performs.
00:05:49.279 --> 00:05:54.959
Because a great marketing manager in one company can be very different to a great marketing manager in another company.
00:05:55.040 --> 00:06:01.279
And that context is super important in terms of the pace, the expectations, the environment, all of that matters.
00:06:01.439 --> 00:06:08.959
So when you simulate the role, you get a much more objective signal, and you can then identify who's actually worth investing time into.
00:06:09.040 --> 00:06:20.639
And that goes on to reduce a lot of the sifting and the screening work the recruiters are doing and ultimately gives the hiring teams much more objective data to then base their decisions on.
00:06:21.040 --> 00:06:26.240
So so skill-based hiring is really becoming popular for those reasons you mentioned.
00:06:26.319 --> 00:06:34.399
And and um, it's almost becoming a must-have, you know, just like like you said, there's just so much, there's a flood of AI-based resumes.
00:06:34.560 --> 00:06:39.600
Um, it's just a changing, you know, dynamic with with the WordPress and what's what's out there.
00:06:39.680 --> 00:06:43.360
So you have to find some solution that works uh for hiring managers, right?
00:06:43.519 --> 00:06:50.000
Um for companies who are interested in in making the shift, where do things usually start to break down?
00:06:50.160 --> 00:06:52.879
What are some of the bottlenecks, the pain points?
00:06:53.199 --> 00:06:57.600
Yeah, I think historically the biggest blocker was always the hiring manager.
00:06:57.759 --> 00:07:00.000
Not because they didn't agree with skills-based hiring.
00:07:00.160 --> 00:07:07.360
Fundamentally, I think, as you've just said, most people you speak to will agree conceptually with skills-based hiring.
00:07:07.519 --> 00:07:10.800
Yeah, the hiring managers are the ones who understand the role.
00:07:10.879 --> 00:07:11.680
They're the experts.
00:07:11.839 --> 00:07:25.040
So previously, if you wanted to build a proper skills-based process, you needed a lot of their time to define the assessment, create the structure, write the questions, and even to go and score every assessment that gets completed by a candidate.
00:07:25.120 --> 00:07:28.000
And in reality, they just don't have the capacity to do that.
00:07:28.160 --> 00:07:34.639
So companies would agree with the idea, but really struggle to go on to implement it into their process.
00:07:34.800 --> 00:07:42.399
I think what's changed now is that all of this can be done so much more efficiently and in some cases instantly with AI.
00:07:42.560 --> 00:07:49.519
So TA teams can now create assessments in minutes without the involvement of a hiring manager purely based on a job description.
00:07:49.680 --> 00:08:00.000
The hiring manager can be involved, don't get me wrong, if they want to come in at a later stage and approve it, they can do that, but they definitely don't need to be as deeply involved in building everything from scratch.
00:08:00.079 --> 00:08:01.839
And that shift has been huge.
00:08:01.920 --> 00:08:08.000
And that's where we've seen a real catalyst to the wider adoption that we're seeing now when it comes to skills-based hiring.
00:08:08.240 --> 00:08:25.839
It also has changed the dynamic between talent and the business because instead of talent just bringing hiring managers this long list of resumes, you're able to bring them a much smaller group of candidates who have already demonstrated they can do the job with an objective work sample of their ability.
00:08:25.920 --> 00:08:29.680
So you're reducing the time that the hiring teams are spending in interviews.
00:08:29.759 --> 00:08:34.320
And on average, with our customers, we see an 80% reduction in interview time.
00:08:34.399 --> 00:08:36.559
Um at Canitech, that's really what we focus on.
00:08:36.639 --> 00:08:49.440
It's about making it very easy to create tailored assessments quickly, whether through our AI builder, where you could just give it a job description, it will create a custom assessment for you, um, and then have everything automatically scored as well.
00:08:49.519 --> 00:08:57.440
So the barrier that used to exist previously, which is the time and the effort required to get started, particularly from the hiring teams, has basically gone now.
00:08:58.080 --> 00:08:59.519
So talk about that timeline.
00:08:59.600 --> 00:09:06.559
Like how is how has things uh the hiring today uh changed from you know the process, let's say even a couple of years ago.
00:09:06.720 --> 00:09:18.720
I know, you know, the the AI revolution has been um you know several years in the in the making now, but um things still change every few months, it seems, with new versions, new iterations, new adoptions.
00:09:18.799 --> 00:09:25.360
Um so yeah, talk about like what's changed for companies how and how they define a strong candidate here in 2026.
00:09:25.919 --> 00:09:28.320
Yeah, I think there's two main shifts.
00:09:28.559 --> 00:09:36.159
The first is that some of the skills that used to be really essential for particular roles are becoming less important, of course, because of AI.
00:09:36.320 --> 00:09:46.399
So to give you an example, previously you may have put a lot of weight on a finance manager having perfect written communication so you can go and communicate with your clients over email.
00:09:46.559 --> 00:09:51.200
But now AI can help them refine any written communication they're sending out.
00:09:51.360 --> 00:09:53.360
So that becomes less essential.
00:09:53.519 --> 00:10:02.480
So the question now is less about can they produce a certain output, but can they use tools to get to the best outcome?
00:10:02.720 --> 00:10:04.080
So that's the first one.
00:10:04.240 --> 00:10:08.320
The second shift is that companies are looking much more holistically at candidates.
00:10:08.399 --> 00:10:10.960
They're no longer just assessing whether someone can do a task.
00:10:11.120 --> 00:10:19.039
Now companies want to evaluate how they approach the work, how they'd explain their thought process, and also how they'd use tools to enhance their work as well.
00:10:19.200 --> 00:10:23.679
And what's really interesting is you can now assess all of that together in one assessment.
00:10:23.840 --> 00:10:31.600
So, for example, if you're hiring a developer, you might ask them to complete a coding challenge, but then explain their thinking and the steps that they took in writing.
00:10:31.759 --> 00:10:37.600
So you can see their logic, you can understand their thought process and see how they got to the outcome they came to.
00:10:37.759 --> 00:10:45.519
And then you could follow that up and ask them to record a short video update as if they were speaking to their manager and giving them a summary of the work they've just performed.
00:10:45.679 --> 00:10:54.000
So when you do something like that, you're no longer relying on signals like years of experience, education, and you're not inferring ability.
00:10:54.080 --> 00:10:56.320
You can actually observe their ability.
00:10:56.480 --> 00:11:05.039
And we've seen this with our own customers who have gone on to widen their candidate pool significantly, and then they go on to actually hire people they wouldn't have considered before.
00:11:05.360 --> 00:11:12.240
To give you an example, one of our customers would only hire people who had a minimum of two years SQL experience.
00:11:12.399 --> 00:11:25.440
Yet when they started working with us, they decided to drop that requirement completely and just send an assessment to anyone who said they had knowledge of SQL, whether that was one year, one and a half years, or even no professional experience at all.
00:11:25.679 --> 00:11:33.600
And what they went on to uh to find out is that they were hiring candidates who they wouldn't even have considered before because of the strict requirement they had.
00:11:33.759 --> 00:11:38.960
So you ultimately end up hiring it from a different pool of candidates than you would have otherwise.
00:11:39.279 --> 00:11:39.759
Interesting.
00:11:39.919 --> 00:11:40.720
Yeah, super interesting.
00:11:40.799 --> 00:11:42.240
That's uh this is really valuable information.
00:11:42.399 --> 00:11:44.240
I hope, I hope listeners are taking note.
00:11:44.399 --> 00:11:55.039
Um, I I I want to uh you know talk about the the AI skills in in the context of um, you know, just the fluency and from the candidates' perspective, let's say.
00:11:55.200 --> 00:12:05.519
Um, you know, I think a lot of times we hear about AI skills in the context of of engineering or maybe some technical work, um, but you know, the conversation's expanding to other areas, of course.
00:12:05.679 --> 00:12:10.399
Basically, every industry is is being impacted by by AI in some way, shape, or form.
00:12:10.559 --> 00:12:25.120
Um, I actually had this conversation yesterday, believe it or not, with a buddy of mine who um is a bit younger than me, but he's he's going back to school to finish his degree, and he's not super confident in in you know his career path or what he wants to do.
00:12:25.279 --> 00:12:30.399
And he actually asked me, he's he said, you know, he's like, What would I recommend he pursue?
00:12:30.480 --> 00:12:32.720
Like, what would I do if I were in his shoes?
00:12:32.960 --> 00:12:51.360
Um, and you know, I basically was just like, I would learn everything there is to learn about AI and the tools available to us, how to fine-tune prompts, how to get specific answers, how to like integrate um, you know, AI, some of these tools into the task you're trying to um, you know, achieve.
00:12:51.519 --> 00:13:02.720
But it got complicated because it's like, well, you know, you're in a higher education uh system that's a little bit, you know, still kind of uh coming to terms with with reality and the um the current landscape.
00:13:02.799 --> 00:13:05.840
And so there's not a lot of like AI specific degrees out there.
00:13:06.000 --> 00:13:20.639
Um, but maybe you can talk to, you know, the AI fluency question of like, as someone with lots of hiring experience, um, do you think, you know, having a lot of AI skills is becoming uh is it a baseline expectation for most jobs now?
00:13:20.799 --> 00:13:26.559
How important is it to be able to use AI, you know, in the context of being employed and employable?
00:13:27.039 --> 00:13:27.919
Yeah, absolutely.
00:13:28.080 --> 00:13:37.039
And you can tell your friend uh from my perspective, they absolutely should take your advice, by the way, and and learn all of these tools because I really do think it's becoming a baseline.
00:13:37.200 --> 00:13:47.759
You know, Guy, our CEO who you met, said something that really stuck with me, which is that hiring someone today who doesn't use AI is a bit like hiring someone in the late 90s who didn't know how to use the internet.
00:13:48.159 --> 00:13:57.759
It's not that these people can't do the job, likely they can, but they're just operating at a completely different level of efficiency when they're using AI versus not using AI.
00:13:57.840 --> 00:14:04.080
And I think what companies are trying to understand now isn't just do you use AI, but how do you use AI?
00:14:04.240 --> 00:14:07.759
Do you use it to think better, move faster, improve your output?
00:14:07.919 --> 00:14:12.879
Or are you just relying on it without much judgment or without much uh critical thinking?
00:14:13.120 --> 00:14:15.840
And that's a distinction that's becoming really important.
00:14:16.240 --> 00:14:24.080
And a lot of our customers, when designing assessments, they want to see how candidates are interacting with AI.
00:14:24.159 --> 00:14:37.039
Yeah, we'll talk about this a bit more, I'm sure, today, but specifically in Candy Tech, we can actually embed ChatGPT directly into assessments, and you can then see the full transcript of the conversation the candidate had as they were working through a task.
00:14:37.120 --> 00:14:43.200
So you can see were they just uh relying too much on uh the output, were they being creative, thinking outside of the box?
00:14:43.279 --> 00:14:47.759
We all know that these tools can hallucinate and sometimes need a bit of steering and direction.
00:14:47.919 --> 00:14:58.799
And that's what companies uh want to see when it comes to AI fluency, that someone understands that and they understand when to push back and how to work in a way that gets to the correct output.
00:14:59.039 --> 00:14:59.519
Gotcha.
00:14:59.600 --> 00:15:03.279
Okay, so I I can go to my friend and say, I was right, you should listen to me.
00:15:03.360 --> 00:15:04.879
I know what I'm talking about.
00:15:05.120 --> 00:15:06.799
Um, no, that's that's good to hear.
00:15:06.960 --> 00:15:13.279
And it's uh it's interesting because I know for his con in his context, it's it there's no formal degree, right, for AI.
00:15:13.360 --> 00:15:18.879
So um I know some professors, some schools are adopting it, but you know, a lot of it has to be self-taught.
00:15:19.039 --> 00:15:31.440
Um, a lot of the institutions haven't really caught up to speed, which is probably what you know you're you're noticing and trying to help out with the skills-based hiring and trying to get them to evaluate skills as opposed to, you know, words on a paper.
00:15:31.600 --> 00:15:36.000
Um, talk about some of the job simulations you utilize in the hiring process.
00:15:36.159 --> 00:15:44.080
How do you uh how do you get them to work um and how how do they help reveal things that a resume or interview might miss?
00:15:44.559 --> 00:15:50.720
Yeah, I think the simplest way to put it is a job simulation lets you see performance, not just potential.
00:15:50.960 --> 00:15:54.240
So a resume is is like someone telling you what they've done.
00:15:54.399 --> 00:15:57.440
An interview is them explaining what they've done.
00:15:57.600 --> 00:15:58.960
A job simulation is different.
00:15:59.039 --> 00:16:00.720
You're actually seeing them do the work.
00:16:00.960 --> 00:16:08.960
For example, if I said to you now, I'm an amazing public speaker, you might believe me, but you would never be 100% confident if I was or I wasn't.
00:16:09.120 --> 00:16:23.279
Whereas if you walked into a room and you saw me on stage delivering a speech to a packed room, very captivated audience, and at the end of the speech, everyone stood up and gave me a standing ovation, you now have objective data on my ability as a public speaker.
00:16:23.519 --> 00:16:36.159
And that's what that's why job simulations tend to be much stronger as a predictor, because you're evaluating someone in a context that's much closer to the real role and you're seeing how they would perform in that situation.
00:16:36.559 --> 00:16:38.960
The other big piece is consistency.
00:16:39.120 --> 00:16:52.960
So when you have every candidate going through the exact same task, under the same conditions, the exact same evaluation framework, you're removing a lot of the variability that you get in interviews where different people might assess different candidates slightly differently.
00:16:53.120 --> 00:17:01.039
There's some data that shows that 47% of the time, different hiring managers would take a different decision on the exact same candidate.
00:17:01.519 --> 00:17:06.079
So you're getting an objective um data point when they go through a simulation.
00:17:06.240 --> 00:17:16.079
And sometimes there's people who don't necessarily stand out on a resume, or maybe they're not the strongest interviewer, but they'll perform exceptionally well when you give them something real to work on.
00:17:16.160 --> 00:17:18.319
And ultimately, that's what you're hiring for.
00:17:18.400 --> 00:17:20.079
You're hiring for ability for skill.
00:17:20.160 --> 00:17:23.039
Um so it's opening up that door for you.
00:17:23.279 --> 00:17:34.799
And what's interesting though, to kind of circle back around to you know, the the AI fluency and the skills required there, it's it's I don't know, I feel like AI is no longer automatically a red flag.
00:17:34.880 --> 00:17:41.200
Um, you know, as we talked, as we talked about in many roles, it's actually, you know, a part of doing the job well.
00:17:41.440 --> 00:17:57.839
Um the the evolution of like scary new technology to powerful new tool, I think is really fascinating from a from a societal and and psychological level, you know, it's like uh the the vision and image I get is is the cavemen when they first kind of witnessed fire.
00:17:57.920 --> 00:18:08.000
They started they were scared of it at first, of course, and then they started to harness it for their benefit, started cooking food, lighting up caves, um uh, you know, a long list of of use cases.
00:18:08.319 --> 00:18:14.319
So when a candidate uses AI in an assessment, what are you really learning from that?
00:18:14.480 --> 00:18:18.559
What does the that reveal about their judgment or their problem solving?
00:18:18.880 --> 00:18:26.160
Yeah, it's actually it's really funny the example that you gave because this was the exact same reaction our customers had when AI was introduced.
00:18:26.319 --> 00:18:34.000
And it's one of the most interesting shifts that we've seen because in the past, using AI in an assessment was considered cheating.
00:18:34.079 --> 00:18:43.920
And actually, a lot of customers came to us saying we need to stop candidates using AI and we want Candidek to build um features so we can detect if they're using AI.
00:18:44.000 --> 00:18:47.440
Uh, but actually that doesn't reflect how work happens anymore.
00:18:47.519 --> 00:18:50.640
You know, in most roles, employers are using AI every day.
00:18:50.799 --> 00:18:55.279
When I walk around our office, everyone's other screen is Chat GPT, and rightly so.
00:18:55.440 --> 00:19:00.559
You know, this is allowing people to move faster, operate at scale, be more efficient, um, as you know.
00:19:00.799 --> 00:19:05.599
So if an assessment is meant to reflect real job performance, then it makes sense to allow it.
00:19:05.680 --> 00:19:08.079
And that's our general recommendation to our customers.
00:19:08.240 --> 00:19:14.559
If they'd be using AI in the role, they should be able to use AI in an assessment, but in a way that it's controlled.
00:19:14.720 --> 00:19:16.559
So you can see how they're using it.
00:19:16.640 --> 00:19:27.839
Um and what that reveals is actually really valuable because you can see how someone works alongside AI, how they use it to enhance their thinking, to move faster, to explore different approaches.
00:19:28.000 --> 00:19:34.079
And if they're just relying on it too heavily without really understanding the output it's it's giving them.
00:19:34.240 --> 00:19:38.880
And that distinction can tell you quite a lot about how they'll go on to perform as an employee.
00:19:39.039 --> 00:19:49.039
At CandyTech, because our assessments are designed to mimic real work, and because you can layer follow-up questions and ask candidates to explain their process, you can actually evaluate both at the same time.
00:19:49.119 --> 00:19:54.480
So you can see their professional knowledge, but then how they would use AI as part of their problem solving.
00:19:54.640 --> 00:20:00.559
So when we have that embedded chat GPT, if you choose to turn it on or not, it's on a case-by-case basis.
00:20:00.799 --> 00:20:06.480
You can go through the full transcript and understand is this person a proficient prompt engineer?
00:20:06.720 --> 00:20:11.759
Did they understand the limitations of AI when they were working with it or not?
00:20:11.920 --> 00:20:14.960
And I think that's becoming a very important skill in itself.
00:20:15.039 --> 00:20:31.920
I heard someone say, I'm definitely paraphrasing, something like one of the most important skills moving forward won't actually be having a particular skill, but it'll be the skill of understanding how to ask the right question to get the output that you need to complete that task.
00:20:32.079 --> 00:20:36.799
Um and I think that's going to be a fundamental shift that we'll see over the next couple of years.
00:20:37.119 --> 00:20:37.759
I can believe it.
00:20:37.839 --> 00:20:38.319
I can believe it.
00:20:38.400 --> 00:20:45.519
Yeah, I mean, it's it's all about asking the right questions, uh, making sure the tool understands the context.
00:20:45.680 --> 00:20:56.319
Um, and um I think it's a good reminder for listeners if you're not like getting into the settings of your your favorite chatbot of choice and like setting up the parameters and the um kind of instructions.
00:20:56.480 --> 00:21:08.720
Um, I think it's a good reminder to go do that and and fine-tune and and tinker with it a little bit and see kind of how it changes the results you get and and the personalities, at least, uh, with your with your chatbot of choice.
00:21:08.880 --> 00:21:11.039
So um that's a good reminder.
00:21:11.119 --> 00:21:16.160
But I also I think the way you're mentioning about you know over-reliance on AI is is interesting.
00:21:16.240 --> 00:21:21.839
Um, there's a fine line, I think, between using AI well and relying on it too heavily, right?
00:21:22.000 --> 00:21:30.000
Um, how do you separate you know smart, effective use of AI from the overdependence that could actually become a problem on the job?
00:21:30.319 --> 00:21:42.720
I think the key is to treat AI as something that can enhance your decision making, but it cannot replace your decision making, uh, especially with hiring, because hiring is still fundamentally about people.
00:21:42.880 --> 00:21:44.240
It's people hiring people.
00:21:44.480 --> 00:21:55.039
AI can be incredibly useful, especially in the context of what we discussed earlier, where you've got this huge volume of applications and companies really struggling to get through some of that noise.
00:21:55.200 --> 00:21:59.759
It will help you prioritize the right candidates and bring consistency to the process, but you can't.
00:22:00.240 --> 00:22:01.440
Outsource judgment.
00:22:01.839 --> 00:22:12.160
For example, at CandyTech, where we use our AI scoring, we make sure to keep it super transparent so you can see why AI has scored something in the way that it has.
00:22:12.240 --> 00:22:15.920
And if you want, you can adjust the scoring mechanics, you can override it.
00:22:16.079 --> 00:22:17.839
So the human is always in control.
00:22:17.920 --> 00:22:19.119
And I think that's really important.
00:22:19.200 --> 00:22:23.920
It should never feel like a black box where we just give away the control to the machine.
00:22:24.160 --> 00:22:35.200
I think the balance will be using AI to guide you so you can make more data-driven decisions and be more efficient, but always keeping the final decision with a human.
00:22:35.519 --> 00:22:36.319
That's good advice.
00:22:36.400 --> 00:22:36.640
Yeah.
00:22:36.720 --> 00:22:38.400
I mean, use it, use it for guidance.
00:22:38.480 --> 00:22:41.759
Don't um let it replace your reasoning.
00:22:41.839 --> 00:23:00.079
Um, because I think that at a certain point, you know, you really stop to you stop developing your critical thinking skills uh by offloading all your you know work to AI where you so you can't even think for yourself, you can't um, you know, answer some some some basic kind of uh tests and scenarios and and skill and and evaluations.
00:23:00.319 --> 00:23:02.559
Um and I think that's just gonna hurt you in the long run.
00:23:02.720 --> 00:23:05.279
So, you know, finding that balance is is really key.
00:23:05.359 --> 00:23:15.920
And um, I think you know, consuming um reading long form content is a good kind of helps your attention span and um helps you to kind of uh absorb information better.
00:23:16.000 --> 00:23:19.440
And there's uh there's a bunch of other you know um things you can do to improve that.
00:23:19.599 --> 00:23:32.000
But um one of my favorite aspects though with the shift to skills-based hiring is just how it uh widens the funnel, I would say, for for talent, uh talented people who don't necessarily look perfect on paper.
00:23:32.160 --> 00:23:41.359
Um have you seen examples where you know better assessments actually helped uncover candidates who may have just been overlooked otherwise because the resume didn't like particularly stand out?
00:23:41.839 --> 00:23:43.200
Yeah, actually all the time.
00:23:43.279 --> 00:24:01.440
And it's it's part of the part of the role that I find the most meaningful because if you actually think about it, with traditional hiring, uh you can't even quantify how many times the best candidate for a role has probably just been filtered out before anyone even speaks to them, just based on where they studied or where they've worked previously or how many years of experience that they had.
00:24:01.680 --> 00:24:06.400
What I love about skills-based hiring is it changes that because it's a meritocracy.
00:24:06.559 --> 00:24:14.480
So you're just giving candidates an opportunity to demonstrate what they can actually do, and then companies can discover people that they probably wouldn't have considered before.
00:24:14.640 --> 00:24:26.400
Similar to the example I gave you earlier with SQL, when you drop the requirement and that rigid um years of experience requirement, you end up hiring people who you just wouldn't even have entertained um when you had that requirement in place.
00:24:26.640 --> 00:24:37.279
And we often hear that those people end up performing incredibly in the role, um, even though they didn't look strong on paper or didn't match the parameters that would have been set up before.
00:24:37.440 --> 00:24:48.240
Um, we hear it a lot when companies open up the funnel just to let anyone take an assessment who obviously has uh the right sort of experience, but just not in that rigid way.
00:24:48.319 --> 00:24:55.359
And then they prioritize who they bring into the next stages of the process purely based on their performance in the assessment.
00:24:55.680 --> 00:25:04.400
Okay, so even if you don't beef up your resume with AI, let's say, maybe you're a Luddite, uh you can still compete with those that do with this approach.
00:25:04.480 --> 00:25:09.039
You can still kind of get behind uh the skills-based uh hiring transition.
00:25:09.279 --> 00:25:12.640
Um, I think that's that's worth worth underscoring right there.
00:25:12.880 --> 00:25:17.519
Now, um, as we know with every organization, you know, they they have their own way of doing things.
00:25:17.680 --> 00:25:28.400
I imagine some teams are are very ready for this kind of evaluation, um, shift and hiring, while others are still more you know attached and you know used to the traditional methods.
00:25:28.559 --> 00:25:34.960
Um, across the different roles you see, which which ones have have been quickest to embrace simulations and AI-based evaluation?
00:25:35.119 --> 00:25:38.160
And then which ones still need more convincing?
00:25:38.480 --> 00:25:41.759
I think early on it was definitely tech and tech roles.
00:25:42.000 --> 00:25:44.720
Engineering teams were already used to things like coding challenges.
00:25:44.799 --> 00:25:52.559
So the idea of evaluating someone through a practical task wasn't new to them, it was just a natural extension of what they'd been doing already.
00:25:52.799 --> 00:25:57.519
What's been interesting over the last few years is how that's expanded across different functions.
00:25:57.599 --> 00:26:02.640
And now we see adoption pretty much across every function in an organization.
00:26:03.039 --> 00:26:11.359
Once companies understand that job simulation doesn't have to be generic, that's when we really see it start to click because you can tailor them to a role very precisely.
00:26:11.440 --> 00:26:21.519
So for developers, of course, it might be coding, but for sales, you could ask someone to build out a pipeline or a forecast and then record a video to explain their thinking or do a pitch to a customer.
00:26:21.680 --> 00:26:28.319
For a customer-facing role, it could be how would you respond to a difficult scenario or how would you handle a certain conversation?
00:26:28.559 --> 00:26:31.920
A marketer could be building out a campaign or a strategy.
00:26:32.160 --> 00:26:39.279
Legal teams could be having to review uh contracts, you could have finance teams working through different models in spreadsheets.
00:26:39.440 --> 00:26:43.119
So that's where we see really broad adoption across industries as well, by the way.
00:26:43.200 --> 00:26:54.640
So we work with really large enterprises, companies like uh Monday.com, Wix, Fiverr, Equinox, Alison Young, but also really fast growing startups, scale-ups, and pretty much everything in between.
00:26:54.799 --> 00:26:58.799
And the use case for each can look completely different depending on the environment.
00:26:58.960 --> 00:27:06.880
And I think the companies that move the fastest are the ones who are taking a more unified approach to the way they evaluate talent.
00:27:06.960 --> 00:27:10.559
Because what happens often is that teams end up with fragmented tools.
00:27:10.640 --> 00:27:18.799
So you've got one for coding, one for cognitive, one for behavioral, but none of them really reflect how the role works as a whole.
00:27:18.960 --> 00:27:30.240
And we're now seeing a shift towards something a bit more consolidated where you can evaluate multiple skills across different roles in a way that's still tailored to every position, but sits within one consistent framework.
00:27:30.400 --> 00:27:37.599
And I think that's where simulations and AI-based evaluation starts to scale across a business, not just within one function.
00:27:37.759 --> 00:27:41.839
So it's an exciting time actually to see how that's being adopted.
00:27:42.160 --> 00:27:43.119
It is an exciting time.
00:27:43.200 --> 00:27:43.359
Yeah.
00:27:43.599 --> 00:27:56.000
And it's nice to hear that, yeah, there's broad adoption, but also your uh platform is is able to kind of scale and adopt uh with the industry you're trying to work with, with the company you're trying to work with in regardless of their industry.
00:27:56.160 --> 00:28:16.880
I I I I for the reason the question of the question, I just think of uh, you know, some of the like the legacy uh old school organizations or or clinics, like I think of banks and like DMVs that are just like using old old school systems that maybe would be reluctant to adopt this kind of newer uh simulation-based hiring process.
00:28:17.039 --> 00:28:22.559
But it sounds like it sounds like they're changing their tune and starting to consider it.
00:28:22.880 --> 00:28:38.160
Actually, just to your point, we have certain banks using the platform and even for their tele role and the the assessment and situation could be hey, a customer's just walked in, they want to make this transaction on behalf of a family member, the family member's not with them.
00:28:38.319 --> 00:28:39.599
What would you do in this scenario?
00:28:39.759 --> 00:28:45.440
And these are things that will come up day-to-day in the role, and then you can see how they would think and approach that situation.
00:28:46.079 --> 00:28:46.960
Yeah, absolutely.
00:28:47.039 --> 00:28:47.359
Uh huh.
00:28:47.519 --> 00:28:53.839
Okay, so there's I'm sure there's a bunch of examples and different um specific uh use cases and you can think of.
00:28:54.000 --> 00:29:04.559
Um uh now, now, Nick, believe it or not, there are other skill-based platforms out there trying to solve some of the same problems you are, um, some of which we've actually had on this very podcast.
00:29:04.640 --> 00:29:08.000
Um, and so I want to give you the chance to make your case here.
00:29:08.160 --> 00:29:13.680
Uh, what's something your competitors are getting wrong that you're determined to do differently?
00:29:14.000 --> 00:29:14.640
I think you're right.
00:29:14.720 --> 00:29:20.559
I think that there's a lot of companies who are looking at skills-based hiring, and I think there's a lot of strong companies in this space.
00:29:20.799 --> 00:29:30.400
But where we've been really focused, as we we kind of spoke about already, but it's building something that reflects how hiring works across an entire organization.
00:29:30.559 --> 00:29:38.960
So most companies aren't just hiring for one type of role, they're hiring developers, salespeople, marketers, analysts, customer-facing roles, and they're doing it all at the same time.
00:29:39.119 --> 00:29:42.880
And those roles require very different ways of evaluating skills.
00:29:43.039 --> 00:29:53.119
So, one of the key things for us has been giving our customers flexibility so they can create assessments that are fully tailored to any role rather than forcing them into a fixed format.
00:29:53.279 --> 00:29:59.440
And that's important because the way you assess a developer should look completely different to how you would assess a lawyer, for example.
00:29:59.599 --> 00:30:04.400
But at the same time, companies still want consistency in the way that they run the process.
00:30:04.640 --> 00:30:08.079
The second piece is taking a more holistic view of candidates.
00:30:08.160 --> 00:30:19.440
So instead of testing one thing in isolation, like uh just coding or just a cognitive abilities test, you can combine different dimensions into one flow through Candy Tech.
00:30:19.599 --> 00:30:26.240
So you can assess technical ability, communication, reasoning, how someone explains their thought process all at the same time.
00:30:26.319 --> 00:30:28.960
And that gives you a much more complete picture.
00:30:29.200 --> 00:30:36.960
And of course, there's AI as well, which is a big part of that, not just in terms of scoring assessments, but how candidates interact with it.
00:30:37.039 --> 00:30:50.400
And we've invested a lot of time in enabling teens to embed AI directly into the assessment so you can evaluate how someone works with these tools, not just what they know about these tools, because that's becoming increasingly more and more important.
00:30:50.559 --> 00:30:59.039
And finally, there's rigor behind it as well because customization is powerful, but only when it's grounded in proper evaluation.
00:30:59.200 --> 00:31:07.759
That's why we have a dedicated test development team that helps ensure that every assessment is structured in a way that's actually predictive and meaningful.
00:31:07.920 --> 00:31:14.559
And that team is uh uh PhD psychometricians who really support our customers to build out validated assessments.
00:31:14.640 --> 00:31:24.400
And I say the combination of the flexibility, the holistic approach, the embedded AI, and the evaluation design is what sets Candy Tech apart.
00:31:24.720 --> 00:31:25.119
Well said.
00:31:25.279 --> 00:31:25.519
All right.
00:31:25.599 --> 00:31:26.799
Yeah, you made a strong case there.
00:31:26.880 --> 00:31:30.240
I think now it's it's you it's time for the competitors to respond.
00:31:30.319 --> 00:31:30.960
It's your turn.
00:31:31.119 --> 00:31:31.839
No, that's great.
00:31:31.920 --> 00:31:40.000
Um, I uh I think that it's also, you know, again, worth worth highlighting or underscoring the importance of having an effective, capable team.
00:31:40.160 --> 00:31:48.880
Um, I mean, as we get lost in all these automations and AI and and, you know, um basically kind of lose that human connection and human element.
00:31:49.039 --> 00:31:58.160
But having a capable team, having a good support team, um, you know, just those good human relationships with your with your clients and um is is really important, right?
00:31:58.400 --> 00:32:01.519
So let's look at uh where things are going from here.
00:32:01.680 --> 00:32:09.920
Um up until this point, we've covered a lot of what's what you guys are able to do, uh, the companies and organizations, industries you're working uh with.
00:32:10.319 --> 00:32:12.319
Um things are moving fast.
00:32:12.480 --> 00:32:14.000
That that goes without saying.
00:32:14.160 --> 00:32:19.920
Um, but at the same time, it's still it still feels like we're pretty early in all of this in subways.
00:32:20.079 --> 00:32:27.200
Where do you think hiring is headed in, let's say, the next few years if these trends keep accelerating like they are?
00:32:27.519 --> 00:32:30.880
Yeah, and I think things are moving at a really rapid pace.
00:32:31.039 --> 00:32:40.160
And most likely we're moving towards a world where hiring is much more focused on how people think and how they adapt rather than just what they've done before.
00:32:40.319 --> 00:32:44.640
I also think in the next couple of years we'll start to see new roles that don't exist today.
00:32:44.799 --> 00:32:59.279
And it's it's hard to say what they'll be just because of the rapid uh change and how fast things are moving, but likely it'll be centered around how humans are able to collaborate with machines in very interesting ways.
00:32:59.359 --> 00:33:15.359
And because of the rapid pace, things like AI fluency, judgment, adaptability are becoming more central to hiring decisions right now because companies are having in mind that if things are changing, the people we hire today need to be able to keep up with that change.
00:33:15.519 --> 00:33:27.039
I also think we'll see companies assess candidates more broadly, not just based on role-specific skills, but to see how they could operate in general, especially in relation to AI.
00:33:27.200 --> 00:33:36.319
We're already seeing with some of our customers where they're thinking about assessments rather than just per role, but they're looking at building assessments across different categories of their business.
00:33:36.480 --> 00:33:43.759
So administrative work, data work, coding work, and then they can understand how candidates perform in different contexts.
00:33:44.000 --> 00:33:51.119
I also think as things continue to move quickly, adaptability will become one of the most important things to evaluate.
00:33:51.440 --> 00:34:00.160
Yeah, it's it's hard to it's hard to predict where things are going, but uh it sounds like you've you've you've done a really good job at predicting um thus far.
00:34:00.319 --> 00:34:04.720
And um I think you know you obviously have an inside look at the at where the where where things are going.
00:34:04.799 --> 00:34:12.880
So um yeah, I'd be curious to have you back too in in like six months to a year to see to see where things have changed from this current point in time.
00:34:13.039 --> 00:34:16.079
Um but I think it's safe to say, you know, AI is not slowing down.
00:34:16.159 --> 00:34:25.519
I think the the need to combat some of the fluff out there and evaluate people for what they actually offer and bring to the table is going to be, you know, more in demand than ever.
00:34:25.679 --> 00:34:28.320
Um maybe adoption picks up, maybe it slows down.
00:34:28.400 --> 00:34:31.199
There are ebbs and flows, some good times, some bad.
00:34:31.280 --> 00:34:35.599
Um, but that's kind of a transition of uh into my next question for you.
00:34:35.679 --> 00:34:39.119
And and it's a it's a way of getting into the more reflective and personal side.
00:34:39.199 --> 00:34:45.519
So bear with me here and feel free to take this question with, you know, from a personal perspective or from a company-wide approach.
00:34:45.679 --> 00:34:56.960
But I'm curious, um, you know, when when times maybe are uncertain, uncertain, like what do you do to help you and your team stay motivated and keep your eye on the prize, uh, so to speak?
00:34:57.360 --> 00:34:58.159
Great question.
00:34:58.400 --> 00:35:05.440
So actually, and and there's been times over the years, you know, I've been in hiring for 16 years, and as you said, it definitely goes in peaks and troughs.
00:35:05.679 --> 00:35:06.880
It's actually really interesting.
00:35:07.039 --> 00:35:11.280
So my my journey into hiring actually uh wasn't an accident.
00:35:11.360 --> 00:35:16.159
A lot of people I speak to who worked in uh recruitment and gone into HR Tech, they they always say I fell into it.
00:35:16.320 --> 00:35:17.519
Um that wasn't the case for me.
00:35:17.679 --> 00:35:22.719
My mum actually worked in recruitment, and growing up, I used to say, uh, you know, people would say, What do you want to be at school?
00:35:22.880 --> 00:35:24.239
I'd say, I want to be in recruitment.
00:35:24.639 --> 00:35:26.400
Um, yeah, and I chose to go and do it.
00:35:26.480 --> 00:35:37.360
And and one of the things that my mum used to say all the time, um that she loved about recruitment that always resonated with me is the work that we do has a huge impact on candidates' lives.
00:35:37.679 --> 00:35:48.960
Switching job is huge, it impacts everything about your life, your finances, who you're spending time with, how you're able to show up in your personal life based on how you feel at work.
00:35:49.199 --> 00:35:57.599
So I always keep very central, it's something we talk a lot about at Candy Tech as well, um, is the work that we're doing the real outcome we're trying to achieve.
00:35:57.679 --> 00:36:00.639
And that's really helping people change their lives in the best way.
00:36:00.719 --> 00:36:07.679
And that's why I love skills-based hiring, as I said earlier, because it it makes hiring much more of a meritocracy.
00:36:07.760 --> 00:36:19.920
Uh, whereas traditional um signals, which although I understand why, why they've been used historically, they can almost limit the best candidate from being able to get the job.
00:36:20.079 --> 00:36:26.880
Um, whereas I think uh skills-based hiring is shifting that without relying so heavily on uh on a past experience and things like that.
00:36:27.039 --> 00:36:28.480
So I think that's a key thing for me.
00:36:28.719 --> 00:36:31.519
I think what we really do ultimately um is meaningful.
00:36:31.679 --> 00:36:37.679
Um I'm sure there's uh lots of candidates who have taken their candy tech assessment, gone on to get a role that they loved.
00:36:37.760 --> 00:36:43.199
And uh if we have a small part in uh changing their lives in that way, that makes me uh extremely happy.
00:36:43.440 --> 00:36:44.159
Yeah, I love that.
00:36:44.239 --> 00:36:44.880
I love that story.
00:36:44.960 --> 00:36:52.320
I love that um that message and uh the the big picture and the ripple effect of of change and the impact you're having is a great takeaway.
00:36:52.400 --> 00:37:04.639
You know, it's it's um something I try to keep in mind too, with having these conversations and um you know the the intended but also unintended kind of consequences of just you know, it makes your day you makes your day brighter.
00:37:04.800 --> 00:37:10.639
I feel better having this talking like with you, for example, and I'm gonna go on and um have a better day as a result of it.
00:37:10.719 --> 00:37:17.840
And what I'm gonna do is to be determined, I suppose, but um it'll only serve to benefit from this interaction.
00:37:17.920 --> 00:37:21.360
And the same can be said about you know what you guys are doing with candy tech.
00:37:21.679 --> 00:37:22.880
So um that's great.
00:37:22.960 --> 00:37:24.639
I love I love that you shared that with us.
00:37:24.800 --> 00:37:25.440
Appreciate it.
00:37:25.599 --> 00:37:31.360
Um now at this point, we have listeners who are interested in and maybe getting contact with you and your team and learning more about candy tech.
00:37:31.679 --> 00:37:32.559
Where would you direct them?
00:37:32.639 --> 00:37:34.159
Where should they go to learn more?
00:37:34.480 --> 00:37:34.960
Awesome.
00:37:35.119 --> 00:37:38.239
So you can go to candytech.io, that's our website.
00:37:38.320 --> 00:37:43.440
So you can find a lot of information about us on there, and there's a way you can contact us through the website directly.
00:37:43.679 --> 00:37:46.880
You can reach out to me on LinkedIn, uh Nick Leslie on LinkedIn.
00:37:47.039 --> 00:37:48.159
Um, draw me a connection.
00:37:48.320 --> 00:37:53.360
I'm always keen to speak to people about hiring generally, skills-based hiring or anything else.
00:37:53.519 --> 00:37:56.159
Um, I love just uh connecting and speaking to people.
00:37:56.320 --> 00:37:59.519
Um yeah, those will be the the two best ways to get in touch.
00:37:59.679 --> 00:38:01.519
So no need to be a stranger.
00:38:02.000 --> 00:38:02.559
Awesome.
00:38:02.719 --> 00:38:03.039
All right.
00:38:03.199 --> 00:38:08.000
Nick Leslie, thank you so much for all your insights and your recommendations that you shared with us.
00:38:08.159 --> 00:38:09.119
I really appreciate your time.
00:38:09.199 --> 00:38:12.239
And uh yeah, I would seriously love to have you back one of these days.
00:38:12.719 --> 00:38:12.960
Thank you.
00:38:13.039 --> 00:38:13.920
I really appreciate your time as well.
00:38:14.000 --> 00:38:15.440
It's been a great conversation.
00:38:15.760 --> 00:38:17.760
Thank you all for listening to the Source Forge podcast.
00:38:17.840 --> 00:38:19.119
I'm your host, Bo Hamilton.
00:38:19.199 --> 00:38:23.440
Make sure to subscribe to stay up to date with all of our upcoming B2B software related podcasts.
00:38:23.599 --> 00:38:25.280
I will talk to you in the next one.
00:00:05.120 --> 00:00:08.000
Hello everyone and welcome to the Source Forge podcast.
00:00:08.080 --> 00:00:09.679
I'm your host, Bo Hamilton.
00:00:09.839 --> 00:00:17.679
Now, for a long time, hiring has leaned heavily on resumes, interviews, and a little bit of, I would say, gut instinct.
00:00:17.760 --> 00:00:21.600
Uh, but in the AI era, that whole equation is starting to change.
00:00:21.679 --> 00:00:27.039
You know, there are a lot of ways candidates are are using AI to refine their application materials these days.
00:00:27.120 --> 00:00:37.200
They can tailor their experience to a job description or even prepare super polished answers in advance, both of which are, you know, understandable from the candidate's perspective.
00:00:37.359 --> 00:00:41.359
You want to stand out from the competition and come across as well as possible.
00:00:41.439 --> 00:00:47.520
Um, but from the hiring perspective, the resume, you know, starts to become a little bit less of a signal.
00:00:47.600 --> 00:00:49.920
Um, it's a weaker signal than it used to be.
00:00:50.000 --> 00:00:57.439
Um, hiring managers have to kind of cut through the AI fluff, so to speak, to find the right candidate with the right know-how for the job.
00:00:57.600 --> 00:01:02.079
And so that is why more companies are shifting towards what's called skills-based hiring.
00:01:02.159 --> 00:01:07.760
Um, and they're asking a more direct question of just can this person actually do the work that they say they can do?
00:01:07.920 --> 00:01:10.239
Um, and so that's where you know our guest of the show comes in.
00:01:10.319 --> 00:01:14.239
Today I'm joined by Nick Leslie, director of customer success at Candy Tech.
00:01:14.480 --> 00:01:29.200
Uh, it's an AI-powered skills assessment platform built around job simulations that help companies assess candidates across a variety of metrics, you know, hard skills, soft skills, leadership, uh tests for behavioral and cognitive traits, and even, of course, AI skills as well.
00:01:29.359 --> 00:01:32.079
That's an increasingly important skill set in today's world.
00:01:32.159 --> 00:01:33.519
So we've got a lot to talk about.
00:01:33.599 --> 00:01:41.760
Uh, we're gonna talk about how AI is impacting the hiring process, what good job simulations actually look like, and then what strong candidates look like in the first place.
00:01:41.920 --> 00:01:44.319
So, again, we've got a lot to cover.
00:01:44.480 --> 00:01:46.560
So, without further ado, let's get right into it.
00:01:46.640 --> 00:01:48.079
Nick, welcome to the podcast.
00:01:48.239 --> 00:01:49.040
Glad you could join us.
00:01:49.200 --> 00:01:49.439
Thank you.
00:01:49.599 --> 00:01:50.640
Appreciate you having me on.
00:01:50.879 --> 00:01:51.519
Absolutely.
00:01:51.680 --> 00:02:02.000
Um, so I actually had the pleasure of interviewing your CEO, Guy Burrell, last year uh to get an understanding of you know the solutions your company is working on to help tackle the, you know, in the hiring space.
00:02:02.159 --> 00:02:09.280
Um, for listeners who might have missed that episode, help us understand the company and kind of the lens you're coming from.
00:02:09.680 --> 00:02:10.159
Yeah, for sure.
00:02:10.319 --> 00:02:12.080
Well, so about it again, thank you for having me on.
00:02:12.159 --> 00:02:14.719
Um, I'm Nick, director of Custom Success at Candy Tech.
00:02:14.960 --> 00:02:23.680
Um, and effectively, what Candy Tech does, we're an all-in-one skills assessment platform that allows you to see how someone would perform a job before they get a job.
00:02:23.919 --> 00:02:31.120
Really, what we believe in is creating simulations that mimic tasks that candidates would have to perform if they were employed in your role.
00:02:31.199 --> 00:02:36.479
So you can get a sense of can they do what you need them to be able to do within your environment?
00:02:36.639 --> 00:02:38.879
And we're extremely holistic in the way that we work.
00:02:38.960 --> 00:02:54.080
So we don't just cover one function of a business and we can run skill-based assessments across any area that you're hiring, whether that's finance, data, marketing, legal, of course, technology, um, which is becoming very essential for people to assess somebody's skills.
00:02:54.240 --> 00:02:58.080
And you can translate assessments into any language on the planet as well.
00:02:58.159 --> 00:03:10.639
So if you're hiring uh really in any geography, you're able to run assessments to get a strong signal if a candidate is worth you investing time into to get into the process and ultimately to get a quality higher.
00:03:10.960 --> 00:03:11.360
Got it.
00:03:11.520 --> 00:03:13.599
Okay, thanks for providing that that overview.
00:03:13.759 --> 00:03:22.000
I think it's uh, you know, what you're doing is more important than ever, you know, because I think people people have been talking about how resumes are just becoming less useful um for a while.
00:03:22.159 --> 00:03:27.439
Now that's been kind of the the conversation I've been hearing amongst friends and family who are looking for jobs.
00:03:27.599 --> 00:03:35.280
Um I know that a resume, let's say I think it's fair to say it was never a perfect measure of whether someone could actually do the job.
00:03:35.360 --> 00:03:55.840
Um, but now with AI, it's e it's even easier for candidates to, you know, do things like we talked about, a polishing resume, uh polishing their experience, using specific language that is tailored to the job description and uses really specific keywords and just really just makes them appear stronger than they actually are on paper, um, than they might be in practice.
00:03:56.000 --> 00:04:01.680
So um, yeah, you can you can check a lot of boxes, look great on paper, but you don't really necessarily have the right candidates.
00:04:01.840 --> 00:04:10.000
So I'm curious, in this AI world, how has your thinking changed on on what you know world uh world beyond resumes actually looks like?
00:04:10.319 --> 00:04:13.120
Yeah, and actually you've you've taken a lot of uh my thoughts.
00:04:13.199 --> 00:04:15.120
Uh you said a lot of my thoughts out loud just Sarah.
00:04:15.199 --> 00:04:17.279
Actually, it's really interesting to hear how you describe that.
00:04:17.360 --> 00:04:19.439
I've been working in hiring now for 16 years.
00:04:19.519 --> 00:04:22.160
So I've seen things shift and evolve a lot.
00:04:22.319 --> 00:04:25.839
And I've always thought, as you just said, the resumes are somewhat imperfect.
00:04:26.000 --> 00:04:34.399
You know, they're self-reported, they rely heavily on someone's past achievements, and fundamentally they just don't tell you how someone will perform in a role.
00:04:34.560 --> 00:04:45.759
But now, actually, with the advancements of AI, my thinking shifted from them being imperfect to being increasingly unreliable, particularly as a primary signal of job performance.
00:04:45.839 --> 00:04:57.360
And this has created a very unique challenge for TA because if we look at the broader environment, as you've just said, right now candidates can tailor their CV to pretty much any role they're applying to instantly.
00:04:57.439 --> 00:04:59.680
And they can do it at scale for hundreds of applications.
00:04:59.839 --> 00:05:03.040
So companies are seeing two things happening at the same time.
00:05:03.199 --> 00:05:06.079
One is that application volume is going through the roof.
00:05:06.240 --> 00:05:10.000
And two, in parallel, a lot of resumes are starting to look very similar.
00:05:10.160 --> 00:05:16.240
I hear constantly from our customers that every candidate or every resume they see basically looks the same.
00:05:16.319 --> 00:05:20.160
So it's really hard to identify who should we focus on and who should we get into process.
00:05:20.319 --> 00:05:22.399
And they're getting hundreds of them for every role.
00:05:22.560 --> 00:05:32.480
So I think that's why the conversation is shifting now from how do we screen resumes better to should resumes even be the thing that we rely on in the first place?
00:05:32.639 --> 00:05:36.319
And that's where skills-based hiring is really coming into the conversation.
00:05:36.480 --> 00:05:42.480
The way we do this at Candy Tech, uh and what we see working really well, is giving candidates something that really mirrors the job.
00:05:42.639 --> 00:05:48.959
So a simulation of the role and the company's environment and then seeing how the candidate performs.
00:05:49.279 --> 00:05:54.959
Because a great marketing manager in one company can be very different to a great marketing manager in another company.
00:05:55.040 --> 00:06:01.279
And that context is super important in terms of the pace, the expectations, the environment, all of that matters.
00:06:01.439 --> 00:06:08.959
So when you simulate the role, you get a much more objective signal, and you can then identify who's actually worth investing time into.
00:06:09.040 --> 00:06:20.639
And that goes on to reduce a lot of the sifting and the screening work the recruiters are doing and ultimately gives the hiring teams much more objective data to then base their decisions on.
00:06:21.040 --> 00:06:26.240
So so skill-based hiring is really becoming popular for those reasons you mentioned.
00:06:26.319 --> 00:06:34.399
And and um, it's almost becoming a must-have, you know, just like like you said, there's just so much, there's a flood of AI-based resumes.
00:06:34.560 --> 00:06:39.600
Um, it's just a changing, you know, dynamic with with the WordPress and what's what's out there.
00:06:39.680 --> 00:06:43.360
So you have to find some solution that works uh for hiring managers, right?
00:06:43.519 --> 00:06:50.000
Um for companies who are interested in in making the shift, where do things usually start to break down?
00:06:50.160 --> 00:06:52.879
What are some of the bottlenecks, the pain points?
00:06:53.199 --> 00:06:57.600
Yeah, I think historically the biggest blocker was always the hiring manager.
00:06:57.759 --> 00:07:00.000
Not because they didn't agree with skills-based hiring.
00:07:00.160 --> 00:07:07.360
Fundamentally, I think, as you've just said, most people you speak to will agree conceptually with skills-based hiring.
00:07:07.519 --> 00:07:10.800
Yeah, the hiring managers are the ones who understand the role.
00:07:10.879 --> 00:07:11.680
They're the experts.
00:07:11.839 --> 00:07:25.040
So previously, if you wanted to build a proper skills-based process, you needed a lot of their time to define the assessment, create the structure, write the questions, and even to go and score every assessment that gets completed by a candidate.
00:07:25.120 --> 00:07:28.000
And in reality, they just don't have the capacity to do that.
00:07:28.160 --> 00:07:34.639
So companies would agree with the idea, but really struggle to go on to implement it into their process.
00:07:34.800 --> 00:07:42.399
I think what's changed now is that all of this can be done so much more efficiently and in some cases instantly with AI.
00:07:42.560 --> 00:07:49.519
So TA teams can now create assessments in minutes without the involvement of a hiring manager purely based on a job description.
00:07:49.680 --> 00:08:00.000
The hiring manager can be involved, don't get me wrong, if they want to come in at a later stage and approve it, they can do that, but they definitely don't need to be as deeply involved in building everything from scratch.
00:08:00.079 --> 00:08:01.839
And that shift has been huge.
00:08:01.920 --> 00:08:08.000
And that's where we've seen a real catalyst to the wider adoption that we're seeing now when it comes to skills-based hiring.
00:08:08.240 --> 00:08:25.839
It also has changed the dynamic between talent and the business because instead of talent just bringing hiring managers this long list of resumes, you're able to bring them a much smaller group of candidates who have already demonstrated they can do the job with an objective work sample of their ability.
00:08:25.920 --> 00:08:29.680
So you're reducing the time that the hiring teams are spending in interviews.
00:08:29.759 --> 00:08:34.320
And on average, with our customers, we see an 80% reduction in interview time.
00:08:34.399 --> 00:08:36.559
Um at Canitech, that's really what we focus on.
00:08:36.639 --> 00:08:49.440
It's about making it very easy to create tailored assessments quickly, whether through our AI builder, where you could just give it a job description, it will create a custom assessment for you, um, and then have everything automatically scored as well.
00:08:49.519 --> 00:08:57.440
So the barrier that used to exist previously, which is the time and the effort required to get started, particularly from the hiring teams, has basically gone now.
00:08:58.080 --> 00:08:59.519
So talk about that timeline.
00:08:59.600 --> 00:09:06.559
Like how is how has things uh the hiring today uh changed from you know the process, let's say even a couple of years ago.
00:09:06.720 --> 00:09:18.720
I know, you know, the the AI revolution has been um you know several years in the in the making now, but um things still change every few months, it seems, with new versions, new iterations, new adoptions.
00:09:18.799 --> 00:09:25.360
Um so yeah, talk about like what's changed for companies how and how they define a strong candidate here in 2026.
00:09:25.919 --> 00:09:28.320
Yeah, I think there's two main shifts.
00:09:28.559 --> 00:09:36.159
The first is that some of the skills that used to be really essential for particular roles are becoming less important, of course, because of AI.
00:09:36.320 --> 00:09:46.399
So to give you an example, previously you may have put a lot of weight on a finance manager having perfect written communication so you can go and communicate with your clients over email.
00:09:46.559 --> 00:09:51.200
But now AI can help them refine any written communication they're sending out.
00:09:51.360 --> 00:09:53.360
So that becomes less essential.
00:09:53.519 --> 00:10:02.480
So the question now is less about can they produce a certain output, but can they use tools to get to the best outcome?
00:10:02.720 --> 00:10:04.080
So that's the first one.
00:10:04.240 --> 00:10:08.320
The second shift is that companies are looking much more holistically at candidates.
00:10:08.399 --> 00:10:10.960
They're no longer just assessing whether someone can do a task.
00:10:11.120 --> 00:10:19.039
Now companies want to evaluate how they approach the work, how they'd explain their thought process, and also how they'd use tools to enhance their work as well.
00:10:19.200 --> 00:10:23.679
And what's really interesting is you can now assess all of that together in one assessment.
00:10:23.840 --> 00:10:31.600
So, for example, if you're hiring a developer, you might ask them to complete a coding challenge, but then explain their thinking and the steps that they took in writing.
00:10:31.759 --> 00:10:37.600
So you can see their logic, you can understand their thought process and see how they got to the outcome they came to.
00:10:37.759 --> 00:10:45.519
And then you could follow that up and ask them to record a short video update as if they were speaking to their manager and giving them a summary of the work they've just performed.
00:10:45.679 --> 00:10:54.000
So when you do something like that, you're no longer relying on signals like years of experience, education, and you're not inferring ability.
00:10:54.080 --> 00:10:56.320
You can actually observe their ability.
00:10:56.480 --> 00:11:05.039
And we've seen this with our own customers who have gone on to widen their candidate pool significantly, and then they go on to actually hire people they wouldn't have considered before.
00:11:05.360 --> 00:11:12.240
To give you an example, one of our customers would only hire people who had a minimum of two years SQL experience.
00:11:12.399 --> 00:11:25.440
Yet when they started working with us, they decided to drop that requirement completely and just send an assessment to anyone who said they had knowledge of SQL, whether that was one year, one and a half years, or even no professional experience at all.
00:11:25.679 --> 00:11:33.600
And what they went on to uh to find out is that they were hiring candidates who they wouldn't even have considered before because of the strict requirement they had.
00:11:33.759 --> 00:11:38.960
So you ultimately end up hiring it from a different pool of candidates than you would have otherwise.
00:11:39.279 --> 00:11:39.759
Interesting.
00:11:39.919 --> 00:11:40.720
Yeah, super interesting.
00:11:40.799 --> 00:11:42.240
That's uh this is really valuable information.
00:11:42.399 --> 00:11:44.240
I hope, I hope listeners are taking note.
00:11:44.399 --> 00:11:55.039
Um, I I I want to uh you know talk about the the AI skills in in the context of um, you know, just the fluency and from the candidates' perspective, let's say.
00:11:55.200 --> 00:12:05.519
Um, you know, I think a lot of times we hear about AI skills in the context of of engineering or maybe some technical work, um, but you know, the conversation's expanding to other areas, of course.
00:12:05.679 --> 00:12:10.399
Basically, every industry is is being impacted by by AI in some way, shape, or form.
00:12:10.559 --> 00:12:25.120
Um, I actually had this conversation yesterday, believe it or not, with a buddy of mine who um is a bit younger than me, but he's he's going back to school to finish his degree, and he's not super confident in in you know his career path or what he wants to do.
00:12:25.279 --> 00:12:30.399
And he actually asked me, he's he said, you know, he's like, What would I recommend he pursue?
00:12:30.480 --> 00:12:32.720
Like, what would I do if I were in his shoes?
00:12:32.960 --> 00:12:51.360
Um, and you know, I basically was just like, I would learn everything there is to learn about AI and the tools available to us, how to fine-tune prompts, how to get specific answers, how to like integrate um, you know, AI, some of these tools into the task you're trying to um, you know, achieve.
00:12:51.519 --> 00:13:02.720
But it got complicated because it's like, well, you know, you're in a higher education uh system that's a little bit, you know, still kind of uh coming to terms with with reality and the um the current landscape.
00:13:02.799 --> 00:13:05.840
And so there's not a lot of like AI specific degrees out there.
00:13:06.000 --> 00:13:20.639
Um, but maybe you can talk to, you know, the AI fluency question of like, as someone with lots of hiring experience, um, do you think, you know, having a lot of AI skills is becoming uh is it a baseline expectation for most jobs now?
00:13:20.799 --> 00:13:26.559
How important is it to be able to use AI, you know, in the context of being employed and employable?
00:13:27.039 --> 00:13:27.919
Yeah, absolutely.
00:13:28.080 --> 00:13:37.039
And you can tell your friend uh from my perspective, they absolutely should take your advice, by the way, and and learn all of these tools because I really do think it's becoming a baseline.
00:13:37.200 --> 00:13:47.759
You know, Guy, our CEO who you met, said something that really stuck with me, which is that hiring someone today who doesn't use AI is a bit like hiring someone in the late 90s who didn't know how to use the internet.
00:13:48.159 --> 00:13:57.759
It's not that these people can't do the job, likely they can, but they're just operating at a completely different level of efficiency when they're using AI versus not using AI.
00:13:57.840 --> 00:14:04.080
And I think what companies are trying to understand now isn't just do you use AI, but how do you use AI?
00:14:04.240 --> 00:14:07.759
Do you use it to think better, move faster, improve your output?
00:14:07.919 --> 00:14:12.879
Or are you just relying on it without much judgment or without much uh critical thinking?
00:14:13.120 --> 00:14:15.840
And that's a distinction that's becoming really important.
00:14:16.240 --> 00:14:24.080
And a lot of our customers, when designing assessments, they want to see how candidates are interacting with AI.
00:14:24.159 --> 00:14:37.039
Yeah, we'll talk about this a bit more, I'm sure, today, but specifically in Candy Tech, we can actually embed ChatGPT directly into assessments, and you can then see the full transcript of the conversation the candidate had as they were working through a task.
00:14:37.120 --> 00:14:43.200
So you can see were they just uh relying too much on uh the output, were they being creative, thinking outside of the box?
00:14:43.279 --> 00:14:47.759
We all know that these tools can hallucinate and sometimes need a bit of steering and direction.
00:14:47.919 --> 00:14:58.799
And that's what companies uh want to see when it comes to AI fluency, that someone understands that and they understand when to push back and how to work in a way that gets to the correct output.
00:14:59.039 --> 00:14:59.519
Gotcha.
00:14:59.600 --> 00:15:03.279
Okay, so I I can go to my friend and say, I was right, you should listen to me.
00:15:03.360 --> 00:15:04.879
I know what I'm talking about.
00:15:05.120 --> 00:15:06.799
Um, no, that's that's good to hear.
00:15:06.960 --> 00:15:13.279
And it's uh it's interesting because I know for his con in his context, it's it there's no formal degree, right, for AI.
00:15:13.360 --> 00:15:18.879
So um I know some professors, some schools are adopting it, but you know, a lot of it has to be self-taught.
00:15:19.039 --> 00:15:31.440
Um, a lot of the institutions haven't really caught up to speed, which is probably what you know you're you're noticing and trying to help out with the skills-based hiring and trying to get them to evaluate skills as opposed to, you know, words on a paper.
00:15:31.600 --> 00:15:36.000
Um, talk about some of the job simulations you utilize in the hiring process.
00:15:36.159 --> 00:15:44.080
How do you uh how do you get them to work um and how how do they help reveal things that a resume or interview might miss?
00:15:44.559 --> 00:15:50.720
Yeah, I think the simplest way to put it is a job simulation lets you see performance, not just potential.
00:15:50.960 --> 00:15:54.240
So a resume is is like someone telling you what they've done.
00:15:54.399 --> 00:15:57.440
An interview is them explaining what they've done.
00:15:57.600 --> 00:15:58.960
A job simulation is different.
00:15:59.039 --> 00:16:00.720
You're actually seeing them do the work.
00:16:00.960 --> 00:16:08.960
For example, if I said to you now, I'm an amazing public speaker, you might believe me, but you would never be 100% confident if I was or I wasn't.
00:16:09.120 --> 00:16:23.279
Whereas if you walked into a room and you saw me on stage delivering a speech to a packed room, very captivated audience, and at the end of the speech, everyone stood up and gave me a standing ovation, you now have objective data on my ability as a public speaker.
00:16:23.519 --> 00:16:36.159
And that's what that's why job simulations tend to be much stronger as a predictor, because you're evaluating someone in a context that's much closer to the real role and you're seeing how they would perform in that situation.
00:16:36.559 --> 00:16:38.960
The other big piece is consistency.
00:16:39.120 --> 00:16:52.960
So when you have every candidate going through the exact same task, under the same conditions, the exact same evaluation framework, you're removing a lot of the variability that you get in interviews where different people might assess different candidates slightly differently.
00:16:53.120 --> 00:17:01.039
There's some data that shows that 47% of the time, different hiring managers would take a different decision on the exact same candidate.
00:17:01.519 --> 00:17:06.079
So you're getting an objective um data point when they go through a simulation.
00:17:06.240 --> 00:17:16.079
And sometimes there's people who don't necessarily stand out on a resume, or maybe they're not the strongest interviewer, but they'll perform exceptionally well when you give them something real to work on.
00:17:16.160 --> 00:17:18.319
And ultimately, that's what you're hiring for.
00:17:18.400 --> 00:17:20.079
You're hiring for ability for skill.
00:17:20.160 --> 00:17:23.039
Um so it's opening up that door for you.
00:17:23.279 --> 00:17:34.799
And what's interesting though, to kind of circle back around to you know, the the AI fluency and the skills required there, it's it's I don't know, I feel like AI is no longer automatically a red flag.
00:17:34.880 --> 00:17:41.200
Um, you know, as we talked, as we talked about in many roles, it's actually, you know, a part of doing the job well.
00:17:41.440 --> 00:17:57.839
Um the the evolution of like scary new technology to powerful new tool, I think is really fascinating from a from a societal and and psychological level, you know, it's like uh the the vision and image I get is is the cavemen when they first kind of witnessed fire.
00:17:57.920 --> 00:18:08.000
They started they were scared of it at first, of course, and then they started to harness it for their benefit, started cooking food, lighting up caves, um uh, you know, a long list of of use cases.
00:18:08.319 --> 00:18:14.319
So when a candidate uses AI in an assessment, what are you really learning from that?
00:18:14.480 --> 00:18:18.559
What does the that reveal about their judgment or their problem solving?
00:18:18.880 --> 00:18:26.160
Yeah, it's actually it's really funny the example that you gave because this was the exact same reaction our customers had when AI was introduced.
00:18:26.319 --> 00:18:34.000
And it's one of the most interesting shifts that we've seen because in the past, using AI in an assessment was considered cheating.
00:18:34.079 --> 00:18:43.920
And actually, a lot of customers came to us saying we need to stop candidates using AI and we want Candidek to build um features so we can detect if they're using AI.
00:18:44.000 --> 00:18:47.440
Uh, but actually that doesn't reflect how work happens anymore.
00:18:47.519 --> 00:18:50.640
You know, in most roles, employers are using AI every day.
00:18:50.799 --> 00:18:55.279
When I walk around our office, everyone's other screen is Chat GPT, and rightly so.
00:18:55.440 --> 00:19:00.559
You know, this is allowing people to move faster, operate at scale, be more efficient, um, as you know.
00:19:00.799 --> 00:19:05.599
So if an assessment is meant to reflect real job performance, then it makes sense to allow it.
00:19:05.680 --> 00:19:08.079
And that's our general recommendation to our customers.
00:19:08.240 --> 00:19:14.559
If they'd be using AI in the role, they should be able to use AI in an assessment, but in a way that it's controlled.
00:19:14.720 --> 00:19:16.559
So you can see how they're using it.
00:19:16.640 --> 00:19:27.839
Um and what that reveals is actually really valuable because you can see how someone works alongside AI, how they use it to enhance their thinking, to move faster, to explore different approaches.
00:19:28.000 --> 00:19:34.079
And if they're just relying on it too heavily without really understanding the output it's it's giving them.
00:19:34.240 --> 00:19:38.880
And that distinction can tell you quite a lot about how they'll go on to perform as an employee.
00:19:39.039 --> 00:19:49.039
At CandyTech, because our assessments are designed to mimic real work, and because you can layer follow-up questions and ask candidates to explain their process, you can actually evaluate both at the same time.
00:19:49.119 --> 00:19:54.480
So you can see their professional knowledge, but then how they would use AI as part of their problem solving.
00:19:54.640 --> 00:20:00.559
So when we have that embedded chat GPT, if you choose to turn it on or not, it's on a case-by-case basis.
00:20:00.799 --> 00:20:06.480
You can go through the full transcript and understand is this person a proficient prompt engineer?
00:20:06.720 --> 00:20:11.759
Did they understand the limitations of AI when they were working with it or not?
00:20:11.920 --> 00:20:14.960
And I think that's becoming a very important skill in itself.
00:20:15.039 --> 00:20:31.920
I heard someone say, I'm definitely paraphrasing, something like one of the most important skills moving forward won't actually be having a particular skill, but it'll be the skill of understanding how to ask the right question to get the output that you need to complete that task.
00:20:32.079 --> 00:20:36.799
Um and I think that's going to be a fundamental shift that we'll see over the next couple of years.
00:20:37.119 --> 00:20:37.759
I can believe it.
00:20:37.839 --> 00:20:38.319
I can believe it.
00:20:38.400 --> 00:20:45.519
Yeah, I mean, it's it's all about asking the right questions, uh, making sure the tool understands the context.
00:20:45.680 --> 00:20:56.319
Um, and um I think it's a good reminder for listeners if you're not like getting into the settings of your your favorite chatbot of choice and like setting up the parameters and the um kind of instructions.
00:20:56.480 --> 00:21:08.720
Um, I think it's a good reminder to go do that and and fine-tune and and tinker with it a little bit and see kind of how it changes the results you get and and the personalities, at least, uh, with your with your chatbot of choice.
00:21:08.880 --> 00:21:11.039
So um that's a good reminder.
00:21:11.119 --> 00:21:16.160
But I also I think the way you're mentioning about you know over-reliance on AI is is interesting.
00:21:16.240 --> 00:21:21.839
Um, there's a fine line, I think, between using AI well and relying on it too heavily, right?
00:21:22.000 --> 00:21:30.000
Um, how do you separate you know smart, effective use of AI from the overdependence that could actually become a problem on the job?
00:21:30.319 --> 00:21:42.720
I think the key is to treat AI as something that can enhance your decision making, but it cannot replace your decision making, uh, especially with hiring, because hiring is still fundamentally about people.
00:21:42.880 --> 00:21:44.240
It's people hiring people.
00:21:44.480 --> 00:21:55.039
AI can be incredibly useful, especially in the context of what we discussed earlier, where you've got this huge volume of applications and companies really struggling to get through some of that noise.
00:21:55.200 --> 00:21:59.759
It will help you prioritize the right candidates and bring consistency to the process, but you can't.
00:22:00.240 --> 00:22:01.440
Outsource judgment.
00:22:01.839 --> 00:22:12.160
For example, at CandyTech, where we use our AI scoring, we make sure to keep it super transparent so you can see why AI has scored something in the way that it has.
00:22:12.240 --> 00:22:15.920
And if you want, you can adjust the scoring mechanics, you can override it.
00:22:16.079 --> 00:22:17.839
So the human is always in control.
00:22:17.920 --> 00:22:19.119
And I think that's really important.
00:22:19.200 --> 00:22:23.920
It should never feel like a black box where we just give away the control to the machine.
00:22:24.160 --> 00:22:35.200
I think the balance will be using AI to guide you so you can make more data-driven decisions and be more efficient, but always keeping the final decision with a human.
00:22:35.519 --> 00:22:36.319
That's good advice.
00:22:36.400 --> 00:22:36.640
Yeah.
00:22:36.720 --> 00:22:38.400
I mean, use it, use it for guidance.
00:22:38.480 --> 00:22:41.759
Don't um let it replace your reasoning.
00:22:41.839 --> 00:23:00.079
Um, because I think that at a certain point, you know, you really stop to you stop developing your critical thinking skills uh by offloading all your you know work to AI where you so you can't even think for yourself, you can't um, you know, answer some some some basic kind of uh tests and scenarios and and skill and and evaluations.
00:23:00.319 --> 00:23:02.559
Um and I think that's just gonna hurt you in the long run.
00:23:02.720 --> 00:23:05.279
So, you know, finding that balance is is really key.
00:23:05.359 --> 00:23:15.920
And um, I think you know, consuming um reading long form content is a good kind of helps your attention span and um helps you to kind of uh absorb information better.
00:23:16.000 --> 00:23:19.440
And there's uh there's a bunch of other you know um things you can do to improve that.
00:23:19.599 --> 00:23:32.000
But um one of my favorite aspects though with the shift to skills-based hiring is just how it uh widens the funnel, I would say, for for talent, uh talented people who don't necessarily look perfect on paper.
00:23:32.160 --> 00:23:41.359
Um have you seen examples where you know better assessments actually helped uncover candidates who may have just been overlooked otherwise because the resume didn't like particularly stand out?
00:23:41.839 --> 00:23:43.200
Yeah, actually all the time.
00:23:43.279 --> 00:24:01.440
And it's it's part of the part of the role that I find the most meaningful because if you actually think about it, with traditional hiring, uh you can't even quantify how many times the best candidate for a role has probably just been filtered out before anyone even speaks to them, just based on where they studied or where they've worked previously or how many years of experience that they had.
00:24:01.680 --> 00:24:06.400
What I love about skills-based hiring is it changes that because it's a meritocracy.
00:24:06.559 --> 00:24:14.480
So you're just giving candidates an opportunity to demonstrate what they can actually do, and then companies can discover people that they probably wouldn't have considered before.
00:24:14.640 --> 00:24:26.400
Similar to the example I gave you earlier with SQL, when you drop the requirement and that rigid um years of experience requirement, you end up hiring people who you just wouldn't even have entertained um when you had that requirement in place.
00:24:26.640 --> 00:24:37.279
And we often hear that those people end up performing incredibly in the role, um, even though they didn't look strong on paper or didn't match the parameters that would have been set up before.
00:24:37.440 --> 00:24:48.240
Um, we hear it a lot when companies open up the funnel just to let anyone take an assessment who obviously has uh the right sort of experience, but just not in that rigid way.
00:24:48.319 --> 00:24:55.359
And then they prioritize who they bring into the next stages of the process purely based on their performance in the assessment.
00:24:55.680 --> 00:25:04.400
Okay, so even if you don't beef up your resume with AI, let's say, maybe you're a Luddite, uh you can still compete with those that do with this approach.
00:25:04.480 --> 00:25:09.039
You can still kind of get behind uh the skills-based uh hiring transition.
00:25:09.279 --> 00:25:12.640
Um, I think that's that's worth worth underscoring right there.
00:25:12.880 --> 00:25:17.519
Now, um, as we know with every organization, you know, they they have their own way of doing things.
00:25:17.680 --> 00:25:28.400
I imagine some teams are are very ready for this kind of evaluation, um, shift and hiring, while others are still more you know attached and you know used to the traditional methods.
00:25:28.559 --> 00:25:34.960
Um, across the different roles you see, which which ones have have been quickest to embrace simulations and AI-based evaluation?
00:25:35.119 --> 00:25:38.160
And then which ones still need more convincing?
00:25:38.480 --> 00:25:41.759
I think early on it was definitely tech and tech roles.
00:25:42.000 --> 00:25:44.720
Engineering teams were already used to things like coding challenges.
00:25:44.799 --> 00:25:52.559
So the idea of evaluating someone through a practical task wasn't new to them, it was just a natural extension of what they'd been doing already.
00:25:52.799 --> 00:25:57.519
What's been interesting over the last few years is how that's expanded across different functions.
00:25:57.599 --> 00:26:02.640
And now we see adoption pretty much across every function in an organization.
00:26:03.039 --> 00:26:11.359
Once companies understand that job simulation doesn't have to be generic, that's when we really see it start to click because you can tailor them to a role very precisely.
00:26:11.440 --> 00:26:21.519
So for developers, of course, it might be coding, but for sales, you could ask someone to build out a pipeline or a forecast and then record a video to explain their thinking or do a pitch to a customer.
00:26:21.680 --> 00:26:28.319
For a customer-facing role, it could be how would you respond to a difficult scenario or how would you handle a certain conversation?
00:26:28.559 --> 00:26:31.920
A marketer could be building out a campaign or a strategy.
00:26:32.160 --> 00:26:39.279
Legal teams could be having to review uh contracts, you could have finance teams working through different models in spreadsheets.
00:26:39.440 --> 00:26:43.119
So that's where we see really broad adoption across industries as well, by the way.
00:26:43.200 --> 00:26:54.640
So we work with really large enterprises, companies like uh Monday.com, Wix, Fiverr, Equinox, Alison Young, but also really fast growing startups, scale-ups, and pretty much everything in between.
00:26:54.799 --> 00:26:58.799
And the use case for each can look completely different depending on the environment.
00:26:58.960 --> 00:27:06.880
And I think the companies that move the fastest are the ones who are taking a more unified approach to the way they evaluate talent.
00:27:06.960 --> 00:27:10.559
Because what happens often is that teams end up with fragmented tools.
00:27:10.640 --> 00:27:18.799
So you've got one for coding, one for cognitive, one for behavioral, but none of them really reflect how the role works as a whole.
00:27:18.960 --> 00:27:30.240
And we're now seeing a shift towards something a bit more consolidated where you can evaluate multiple skills across different roles in a way that's still tailored to every position, but sits within one consistent framework.
00:27:30.400 --> 00:27:37.599
And I think that's where simulations and AI-based evaluation starts to scale across a business, not just within one function.
00:27:37.759 --> 00:27:41.839
So it's an exciting time actually to see how that's being adopted.
00:27:42.160 --> 00:27:43.119
It is an exciting time.
00:27:43.200 --> 00:27:43.359
Yeah.
00:27:43.599 --> 00:27:56.000
And it's nice to hear that, yeah, there's broad adoption, but also your uh platform is is able to kind of scale and adopt uh with the industry you're trying to work with, with the company you're trying to work with in regardless of their industry.
00:27:56.160 --> 00:28:16.880
I I I I for the reason the question of the question, I just think of uh, you know, some of the like the legacy uh old school organizations or or clinics, like I think of banks and like DMVs that are just like using old old school systems that maybe would be reluctant to adopt this kind of newer uh simulation-based hiring process.
00:28:17.039 --> 00:28:22.559
But it sounds like it sounds like they're changing their tune and starting to consider it.
00:28:22.880 --> 00:28:38.160
Actually, just to your point, we have certain banks using the platform and even for their tele role and the the assessment and situation could be hey, a customer's just walked in, they want to make this transaction on behalf of a family member, the family member's not with them.
00:28:38.319 --> 00:28:39.599
What would you do in this scenario?
00:28:39.759 --> 00:28:45.440
And these are things that will come up day-to-day in the role, and then you can see how they would think and approach that situation.
00:28:46.079 --> 00:28:46.960
Yeah, absolutely.
00:28:47.039 --> 00:28:47.359
Uh huh.
00:28:47.519 --> 00:28:53.839
Okay, so there's I'm sure there's a bunch of examples and different um specific uh use cases and you can think of.
00:28:54.000 --> 00:29:04.559
Um uh now, now, Nick, believe it or not, there are other skill-based platforms out there trying to solve some of the same problems you are, um, some of which we've actually had on this very podcast.
00:29:04.640 --> 00:29:08.000
Um, and so I want to give you the chance to make your case here.
00:29:08.160 --> 00:29:13.680
Uh, what's something your competitors are getting wrong that you're determined to do differently?
00:29:14.000 --> 00:29:14.640
I think you're right.
00:29:14.720 --> 00:29:20.559
I think that there's a lot of companies who are looking at skills-based hiring, and I think there's a lot of strong companies in this space.
00:29:20.799 --> 00:29:30.400
But where we've been really focused, as we we kind of spoke about already, but it's building something that reflects how hiring works across an entire organization.
00:29:30.559 --> 00:29:38.960
So most companies aren't just hiring for one type of role, they're hiring developers, salespeople, marketers, analysts, customer-facing roles, and they're doing it all at the same time.
00:29:39.119 --> 00:29:42.880
And those roles require very different ways of evaluating skills.
00:29:43.039 --> 00:29:53.119
So, one of the key things for us has been giving our customers flexibility so they can create assessments that are fully tailored to any role rather than forcing them into a fixed format.
00:29:53.279 --> 00:29:59.440
And that's important because the way you assess a developer should look completely different to how you would assess a lawyer, for example.
00:29:59.599 --> 00:30:04.400
But at the same time, companies still want consistency in the way that they run the process.
00:30:04.640 --> 00:30:08.079
The second piece is taking a more holistic view of candidates.
00:30:08.160 --> 00:30:19.440
So instead of testing one thing in isolation, like uh just coding or just a cognitive abilities test, you can combine different dimensions into one flow through Candy Tech.
00:30:19.599 --> 00:30:26.240
So you can assess technical ability, communication, reasoning, how someone explains their thought process all at the same time.
00:30:26.319 --> 00:30:28.960
And that gives you a much more complete picture.
00:30:29.200 --> 00:30:36.960
And of course, there's AI as well, which is a big part of that, not just in terms of scoring assessments, but how candidates interact with it.
00:30:37.039 --> 00:30:50.400
And we've invested a lot of time in enabling teens to embed AI directly into the assessment so you can evaluate how someone works with these tools, not just what they know about these tools, because that's becoming increasingly more and more important.
00:30:50.559 --> 00:30:59.039
And finally, there's rigor behind it as well because customization is powerful, but only when it's grounded in proper evaluation.
00:30:59.200 --> 00:31:07.759
That's why we have a dedicated test development team that helps ensure that every assessment is structured in a way that's actually predictive and meaningful.
00:31:07.920 --> 00:31:14.559
And that team is uh uh PhD psychometricians who really support our customers to build out validated assessments.
00:31:14.640 --> 00:31:24.400
And I say the combination of the flexibility, the holistic approach, the embedded AI, and the evaluation design is what sets Candy Tech apart.
00:31:24.720 --> 00:31:25.119
Well said.
00:31:25.279 --> 00:31:25.519
All right.
00:31:25.599 --> 00:31:26.799
Yeah, you made a strong case there.
00:31:26.880 --> 00:31:30.240
I think now it's it's you it's time for the competitors to respond.
00:31:30.319 --> 00:31:30.960
It's your turn.
00:31:31.119 --> 00:31:31.839
No, that's great.
00:31:31.920 --> 00:31:40.000
Um, I uh I think that it's also, you know, again, worth worth highlighting or underscoring the importance of having an effective, capable team.
00:31:40.160 --> 00:31:48.880
Um, I mean, as we get lost in all these automations and AI and and, you know, um basically kind of lose that human connection and human element.
00:31:49.039 --> 00:31:58.160
But having a capable team, having a good support team, um, you know, just those good human relationships with your with your clients and um is is really important, right?
00:31:58.400 --> 00:32:01.519
So let's look at uh where things are going from here.
00:32:01.680 --> 00:32:09.920
Um up until this point, we've covered a lot of what's what you guys are able to do, uh, the companies and organizations, industries you're working uh with.
00:32:10.319 --> 00:32:12.319
Um things are moving fast.
00:32:12.480 --> 00:32:14.000
That that goes without saying.
00:32:14.160 --> 00:32:19.920
Um, but at the same time, it's still it still feels like we're pretty early in all of this in subways.
00:32:20.079 --> 00:32:27.200
Where do you think hiring is headed in, let's say, the next few years if these trends keep accelerating like they are?
00:32:27.519 --> 00:32:30.880
Yeah, and I think things are moving at a really rapid pace.
00:32:31.039 --> 00:32:40.160
And most likely we're moving towards a world where hiring is much more focused on how people think and how they adapt rather than just what they've done before.
00:32:40.319 --> 00:32:44.640
I also think in the next couple of years we'll start to see new roles that don't exist today.
00:32:44.799 --> 00:32:59.279
And it's it's hard to say what they'll be just because of the rapid uh change and how fast things are moving, but likely it'll be centered around how humans are able to collaborate with machines in very interesting ways.
00:32:59.359 --> 00:33:15.359
And because of the rapid pace, things like AI fluency, judgment, adaptability are becoming more central to hiring decisions right now because companies are having in mind that if things are changing, the people we hire today need to be able to keep up with that change.
00:33:15.519 --> 00:33:27.039
I also think we'll see companies assess candidates more broadly, not just based on role-specific skills, but to see how they could operate in general, especially in relation to AI.
00:33:27.200 --> 00:33:36.319
We're already seeing with some of our customers where they're thinking about assessments rather than just per role, but they're looking at building assessments across different categories of their business.
00:33:36.480 --> 00:33:43.759
So administrative work, data work, coding work, and then they can understand how candidates perform in different contexts.
00:33:44.000 --> 00:33:51.119
I also think as things continue to move quickly, adaptability will become one of the most important things to evaluate.
00:33:51.440 --> 00:34:00.160
Yeah, it's it's hard to it's hard to predict where things are going, but uh it sounds like you've you've you've done a really good job at predicting um thus far.
00:34:00.319 --> 00:34:04.720
And um I think you know you obviously have an inside look at the at where the where where things are going.
00:34:04.799 --> 00:34:12.880
So um yeah, I'd be curious to have you back too in in like six months to a year to see to see where things have changed from this current point in time.
00:34:13.039 --> 00:34:16.079
Um but I think it's safe to say, you know, AI is not slowing down.
00:34:16.159 --> 00:34:25.519
I think the the need to combat some of the fluff out there and evaluate people for what they actually offer and bring to the table is going to be, you know, more in demand than ever.
00:34:25.679 --> 00:34:28.320
Um maybe adoption picks up, maybe it slows down.
00:34:28.400 --> 00:34:31.199
There are ebbs and flows, some good times, some bad.
00:34:31.280 --> 00:34:35.599
Um, but that's kind of a transition of uh into my next question for you.
00:34:35.679 --> 00:34:39.119
And and it's a it's a way of getting into the more reflective and personal side.
00:34:39.199 --> 00:34:45.519
So bear with me here and feel free to take this question with, you know, from a personal perspective or from a company-wide approach.
00:34:45.679 --> 00:34:56.960
But I'm curious, um, you know, when when times maybe are uncertain, uncertain, like what do you do to help you and your team stay motivated and keep your eye on the prize, uh, so to speak?
00:34:57.360 --> 00:34:58.159
Great question.
00:34:58.400 --> 00:35:05.440
So actually, and and there's been times over the years, you know, I've been in hiring for 16 years, and as you said, it definitely goes in peaks and troughs.
00:35:05.679 --> 00:35:06.880
It's actually really interesting.
00:35:07.039 --> 00:35:11.280
So my my journey into hiring actually uh wasn't an accident.
00:35:11.360 --> 00:35:16.159
A lot of people I speak to who worked in uh recruitment and gone into HR Tech, they they always say I fell into it.
00:35:16.320 --> 00:35:17.519
Um that wasn't the case for me.
00:35:17.679 --> 00:35:22.719
My mum actually worked in recruitment, and growing up, I used to say, uh, you know, people would say, What do you want to be at school?
00:35:22.880 --> 00:35:24.239
I'd say, I want to be in recruitment.
00:35:24.639 --> 00:35:26.400
Um, yeah, and I chose to go and do it.
00:35:26.480 --> 00:35:37.360
And and one of the things that my mum used to say all the time, um that she loved about recruitment that always resonated with me is the work that we do has a huge impact on candidates' lives.
00:35:37.679 --> 00:35:48.960
Switching job is huge, it impacts everything about your life, your finances, who you're spending time with, how you're able to show up in your personal life based on how you feel at work.
00:35:49.199 --> 00:35:57.599
So I always keep very central, it's something we talk a lot about at Candy Tech as well, um, is the work that we're doing the real outcome we're trying to achieve.
00:35:57.679 --> 00:36:00.639
And that's really helping people change their lives in the best way.
00:36:00.719 --> 00:36:07.679
And that's why I love skills-based hiring, as I said earlier, because it it makes hiring much more of a meritocracy.
00:36:07.760 --> 00:36:19.920
Uh, whereas traditional um signals, which although I understand why, why they've been used historically, they can almost limit the best candidate from being able to get the job.
00:36:20.079 --> 00:36:26.880
Um, whereas I think uh skills-based hiring is shifting that without relying so heavily on uh on a past experience and things like that.
00:36:27.039 --> 00:36:28.480
So I think that's a key thing for me.
00:36:28.719 --> 00:36:31.519
I think what we really do ultimately um is meaningful.
00:36:31.679 --> 00:36:37.679
Um I'm sure there's uh lots of candidates who have taken their candy tech assessment, gone on to get a role that they loved.
00:36:37.760 --> 00:36:43.199
And uh if we have a small part in uh changing their lives in that way, that makes me uh extremely happy.
00:36:43.440 --> 00:36:44.159
Yeah, I love that.
00:36:44.239 --> 00:36:44.880
I love that story.
00:36:44.960 --> 00:36:52.320
I love that um that message and uh the the big picture and the ripple effect of of change and the impact you're having is a great takeaway.
00:36:52.400 --> 00:37:04.639
You know, it's it's um something I try to keep in mind too, with having these conversations and um you know the the intended but also unintended kind of consequences of just you know, it makes your day you makes your day brighter.
00:37:04.800 --> 00:37:10.639
I feel better having this talking like with you, for example, and I'm gonna go on and um have a better day as a result of it.
00:37:10.719 --> 00:37:17.840
And what I'm gonna do is to be determined, I suppose, but um it'll only serve to benefit from this interaction.
00:37:17.920 --> 00:37:21.360
And the same can be said about you know what you guys are doing with candy tech.
00:37:21.679 --> 00:37:22.880
So um that's great.
00:37:22.960 --> 00:37:24.639
I love I love that you shared that with us.
00:37:24.800 --> 00:37:25.440
Appreciate it.
00:37:25.599 --> 00:37:31.360
Um now at this point, we have listeners who are interested in and maybe getting contact with you and your team and learning more about candy tech.
00:37:31.679 --> 00:37:32.559
Where would you direct them?
00:37:32.639 --> 00:37:34.159
Where should they go to learn more?
00:37:34.480 --> 00:37:34.960
Awesome.
00:37:35.119 --> 00:37:38.239
So you can go to candytech.io, that's our website.
00:37:38.320 --> 00:37:43.440
So you can find a lot of information about us on there, and there's a way you can contact us through the website directly.
00:37:43.679 --> 00:37:46.880
You can reach out to me on LinkedIn, uh Nick Leslie on LinkedIn.
00:37:47.039 --> 00:37:48.159
Um, draw me a connection.
00:37:48.320 --> 00:37:53.360
I'm always keen to speak to people about hiring generally, skills-based hiring or anything else.
00:37:53.519 --> 00:37:56.159
Um, I love just uh connecting and speaking to people.
00:37:56.320 --> 00:37:59.519
Um yeah, those will be the the two best ways to get in touch.
00:37:59.679 --> 00:38:01.519
So no need to be a stranger.
00:38:02.000 --> 00:38:02.559
Awesome.
00:38:02.719 --> 00:38:03.039
All right.
00:38:03.199 --> 00:38:08.000
Nick Leslie, thank you so much for all your insights and your recommendations that you shared with us.
00:38:08.159 --> 00:38:09.119
I really appreciate your time.
00:38:09.199 --> 00:38:12.239
And uh yeah, I would seriously love to have you back one of these days.
00:38:12.719 --> 00:38:12.960
Thank you.
00:38:13.039 --> 00:38:13.920
I really appreciate your time as well.
00:38:14.000 --> 00:38:15.440
It's been a great conversation.
00:38:15.760 --> 00:38:17.760
Thank you all for listening to the Source Forge podcast.
00:38:17.840 --> 00:38:19.119
I'm your host, Bo Hamilton.
00:38:19.199 --> 00:38:23.440
Make sure to subscribe to stay up to date with all of our upcoming B2B software related podcasts.
00:38:23.599 --> 00:38:25.280
I will talk to you in the next one.
