Listen Notes: o melhor motor de busca de podcasts
From Full Stack To VP: Building Platforms That Survive Traffic Spikes

Transcrição completa de From Full Stack To VP: Building Platforms That Survive Traffic Spikes

TRANSCRIÇÃO 🔗

WEBVTT 00:00:00.000 --> 00:00:00.400 Yalla? 00:00:00.800 --> 00:00:01.120 Yalla. 00:00:16.960 --> 00:00:34.240 Wara kill mi nashha, fiem handis, beshuf is yagairo ma bishufa, wudafna lyom, ha wahat abbraz lasme yal jamait bened albar mäjye, waro yal kiyadi, fi akbar share kitanibil manta, eja sharikna tajp tulafa wiye, wo we ea ye, baidan anarasmi yet. 00:00:34.320 --> 00:00:37.439 Munrahebil VP of engineering lazid, states Hosamin Marzuk. 00:00:37.520 --> 00:00:38.159 Yeah wah. 00:00:38.399 --> 00:00:43.039 Nahna pleasure ktier and we looking forward an amazing uh episode. 00:00:43.280 --> 00:00:43.840 Inshallah. 00:00:44.000 --> 00:00:53.920 Al Abul Mabalish Bus Maul ka VP of Engineering, Adiat Lishray to Heke, Mafishi, wireless mics, be capacity, be uh capability ali. 00:00:54.560 --> 00:01:06.400 Shef any uh I haven't worked in audio specifically, but say any uh to that note my career back in 2014 um as a full stack engineer. 00:01:06.640 --> 00:01:15.519 When I quote full stack engineer, it's not that uh the the common sense uh common any conception of uh full stack engineer where you do back end and front end. 00:01:15.680 --> 00:01:30.480 Uh I was working for uh software RD house, not software, an RD house, uh and we were doing uh all sorts of software and hardware at the same time. 00:01:31.519 --> 00:01:39.040 One of our clients can uh Yatuk uh Yatuk and Arabic coffee maker, gaha Swediyya at the time. 00:01:45.439 --> 00:01:49.599 So we automatic coffee maker, Lil Gahwal Swadiya. 00:01:50.159 --> 00:02:05.840 Uh we were designing chips, sending it to China for fabrication, chips, and then we we flashed hardware uh software are uh we were doing that and a few other products and I worked on hardware. 00:02:06.400 --> 00:02:13.840 Uh Bessi and I learned I did the hardware, software, and then back uh back end well front end. 00:02:14.080 --> 00:02:19.840 And I learned that I really hate software hardware and I don't want to do hardware ever again. 00:02:20.000 --> 00:02:24.080 Uh any especially that debugging hardware is uh any hell. 00:02:24.319 --> 00:02:25.039 It's a hassle. 00:02:25.280 --> 00:02:33.039 Yeah, and you you you you start by checking the chip is actually working and your soldering is correct. 00:02:34.159 --> 00:02:38.159 And then you check the software, and then turns out the software has a bug. 00:02:38.319 --> 00:02:40.240 You fix the bug, but still it's not working. 00:02:40.560 --> 00:02:44.000 You go back, turns out your soldering is is is is incorrect. 00:02:44.080 --> 00:02:46.719 And then um no, I'm good. 00:02:46.960 --> 00:02:54.080 Yeah, we'll skip it like in time, VP of Engineering visit. 00:02:54.479 --> 00:03:01.599 Okay, um, so I started Gabal uh Tabelur, I started uh learning to develop uh mobile apps. 00:03:01.840 --> 00:03:07.680 But it be iOS uh at the time can programming language I was objective-c. 00:03:07.840 --> 00:03:20.479 If I started to learn program uh objective-c programming, uh that gave me the opportunity to meet the founder uh Tabelwur Zadbakari who he he was like, okay, I'll work with us. 00:03:20.800 --> 00:03:24.560 Um and I did nothing related to mobile there. 00:03:24.719 --> 00:03:29.120 Uh I don't know why he was so interested, but was a starting point. 00:03:29.360 --> 00:03:31.039 A chef potential, obviously. 00:03:33.919 --> 00:03:52.479 And then um uh there was any a roller coaster of things that I did, and he at the same time uh I started uh uh a uh um a cloud services, cloud automation company, uh Kansas Mahamad Cloud. 00:03:52.639 --> 00:03:56.879 And um at the time can't be Bernamage akoumisma Bernamage Badly Hadden. 00:03:57.599 --> 00:04:02.960 It was a business incubator, uh disease. 00:04:05.599 --> 00:04:07.599 Uh we were incubated there. 00:04:07.759 --> 00:04:13.120 Um what we were doing is uh offering remote development environment developers. 00:04:13.360 --> 00:04:13.520 Okay. 00:04:13.840 --> 00:04:16.800 Uh Alfa Sotash. 00:04:16.959 --> 00:04:28.560 Um and it was still early days, it was uh still early days, and who would invest, let alone usel products, invest in a uh cloud automation product. 00:04:28.800 --> 00:04:34.560 Um, and that was part of the difficulty that we went through for startup funding. 00:04:34.959 --> 00:04:35.519 Funding. 00:04:35.680 --> 00:04:42.319 Um and okay, Al Hain, we have the product, how would we convince people to use it? 00:04:42.639 --> 00:04:52.959 Um we had this two-fold problem and uh investors are not so mature to uh uh to get the product, so they're not gonna invest unless there are customers. 00:04:53.199 --> 00:05:06.480 If we customers, uh I need custom I need to convince them to use it who they think in uh at some really soon at some point I'll go bankrupt and they will not have uh their development environments available. 00:05:08.959 --> 00:05:10.240 It went back and forth. 00:05:10.319 --> 00:05:17.839 Uh I think by mid uh 2016 uh we got an offer in Murni uh to get acquired. 00:05:17.920 --> 00:05:20.399 So we were acquired by Murni and then we joined Murni. 00:05:20.560 --> 00:05:42.800 Uh Murni, for those who don't know what Murni is, um Murni Bidet uh as uh khidmet um if you heard of AAA in the US um Tifta Hatbir, if you any if you could always ask for any service on tires, uh if it's siar and so on. 00:05:43.040 --> 00:05:51.439 Uh and over time and they grew to a portfolio of products that anything has to do with the word car and home. 00:06:14.959 --> 00:06:16.160 They sell it four parts. 00:06:16.319 --> 00:06:17.439 Uh and everything in between. 00:06:17.519 --> 00:06:28.800 They work with Masher Kata'min uh uh umala, they salhona, but they have all of this portfolio. 00:06:29.360 --> 00:06:31.360 This is recent, uh, where they got to this. 00:06:32.240 --> 00:06:34.000 Uh I was there early days, yeah. 00:06:34.240 --> 00:06:37.680 I was considered uh founding uh engineers at uh at Murni. 00:06:38.079 --> 00:06:51.759 Uh at the time when I joined,
Load full transcript
there was um uh the app was built externally outsourced, yeah, but uh CTO was responsible for backend. 00:06:52.240 --> 00:07:05.519 My mandate was simple come fix the the situation with mobile apps, meaning hire an in-house development team and develop rewrite the apps, uh, and then uh you're free to go. 00:07:05.920 --> 00:07:06.720 Good deal. 00:07:07.120 --> 00:07:07.839 Very good deal. 00:07:08.240 --> 00:07:16.800 Yeah, uh it was in this way hectic because we want we I I needed to hire because I had six months before the end of the contract with the outsourcing. 00:07:17.279 --> 00:07:19.839 Very limited time, yes, to write the new apps. 00:07:20.079 --> 00:07:35.360 Uh we were both writing code and hiring at the same time, and I was uh interviews with uh after interviews with the um with the CTO, uh trying to find the perfect TNE fit for what we wanted. 00:07:35.600 --> 00:07:40.959 Uh and surprisingly, can Assal Nanal get Hadistul allaw us then uh hadith al Android. 00:07:41.199 --> 00:07:41.519 Oof. 00:07:41.920 --> 00:07:42.480 Yeah. 00:07:42.800 --> 00:07:50.560 Uh and uh I at the time I wrote the majority, oh very big portion of the Android app. 00:07:50.879 --> 00:07:53.519 And uh I'm not proud of what I did. 00:07:54.000 --> 00:07:59.199 We we uh we got it there, we released it, but there's a bug, but a lot of bugs. 00:07:59.439 --> 00:08:09.360 Um and feel industry Limorni can uh most of the providers, and you're talking about 2016 iPhones were still expensive, Android phones were cheaper. 00:08:09.759 --> 00:08:16.800 Uh yeah, but you can find uh uh Galaxy Fold for uh Alpha 2 or something like that. 00:08:16.959 --> 00:08:27.120 But um the scannel challenge in uh uh Android the cheap Android phone phones were really cheap, yeah, very bad hardware. 00:08:27.360 --> 00:08:34.320 Uh majority of the providers that we needed uh to track all the time were using cheap Android phones. 00:08:34.960 --> 00:08:38.480 We needed to really optimize the Android app, uh provider app. 00:08:38.559 --> 00:08:44.399 We had four apps, two provider and two customer, uh one uh uh two for each platform. 00:08:44.480 --> 00:08:54.080 Uh but we needed to really optimize and tune that app so that it doesn't uh upuse battery and uh to actually get us accurate locations. 00:08:54.240 --> 00:08:58.960 Then uh uh when we track the trips, you can see we we used to call them helicopter trips. 00:08:59.120 --> 00:09:04.960 Then uh like what how is this a trip? 00:09:05.279 --> 00:09:08.559 Uh it was a really cool experience. 00:09:08.639 --> 00:09:11.679 Yes, I was there full time for seven months. 00:09:11.919 --> 00:09:26.159 Uh the amount of things I had I got to see, dealing with people, especially uh they're different kinds of customers, they're not uh tech literate, they designing the UIs. 00:09:26.320 --> 00:09:29.279 We had to big like big buttons with very bright colours. 00:09:33.440 --> 00:09:35.360 Very big red button, yes. 00:09:35.919 --> 00:09:39.120 Uh it was a different experience. 00:09:40.159 --> 00:09:43.200 The time uh back then, yeah, it was it was still early days, man. 00:09:43.600 --> 00:09:45.519 Um pre-COVID, yeah. 00:09:45.600 --> 00:09:53.679 I mean pre-COVID, uh any post-COVID, yes, a lot of people got to to use uh online services more. 00:09:53.840 --> 00:09:59.279 But pre-COVID, 2016, very early days of the transformation guide service. 00:10:00.080 --> 00:10:00.559 It's very different. 00:10:00.639 --> 00:10:13.679 So I I I always applaud uh uh applaud the um uh Salman for what he did, ma'am or uh yeah, seven months uh I moved to a few other things and then okay. 00:10:13.759 --> 00:10:14.960 How did you get to Zid? 00:10:15.279 --> 00:10:17.919 2018, uh June 2018. 00:10:18.559 --> 00:10:20.559 I knew Mazen long before that. 00:10:20.720 --> 00:10:22.720 Uh best for 2018. 00:10:22.799 --> 00:10:38.879 Can uh um what Mazen was asking for was uh you have a lot of good experience with platforms and holding platforms stable, and Zid is at a stage where we have problems with traffic spikes and uh deployments. 00:10:38.960 --> 00:10:41.519 When we deploy something, the platform goes down. 00:10:41.759 --> 00:10:46.639 When uh a merchant does uh a big campaign, the platform goes down. 00:10:46.879 --> 00:10:52.639 Uh today we're uh he said that today we're at an F, gets us uh get us to S. 00:10:52.960 --> 00:10:53.039 Okay. 00:10:53.200 --> 00:10:53.919 And then you're good to go. 00:10:54.000 --> 00:10:59.279 Lena I was like, I'm not sure if I want to work here or we're still in that dude. 00:10:59.679 --> 00:11:01.120 So he was like, halas. 00:11:01.600 --> 00:11:04.000 Uh fix this and then you're good to go. 00:11:04.240 --> 00:11:08.879 Uh and you see how I I negotiated with partners on the very small scopes. 00:11:09.679 --> 00:11:13.519 Uh but yeah, um, I joined the team. 00:11:13.679 --> 00:11:19.120 Um I worked with on that, migrated the platform to AWS, containerized the platform. 00:11:20.960 --> 00:11:22.399 Uh I was a tech lead. 00:11:22.879 --> 00:11:26.960 Um and we have a distinction between tech lead and a team lead. 00:11:27.039 --> 00:11:29.919 I was not leading anyone, I was leading the project itself. 00:11:31.360 --> 00:11:36.159 Um so my role was just to fix these issues. 00:11:36.399 --> 00:11:38.559 Uh I was the OG DevOps guy. 00:11:38.720 --> 00:11:38.879 Okay. 00:11:39.440 --> 00:11:51.440 Um that took about five or six months, and then in June, uh sorry, in December, uh we renegotiated. 00:11:53.200 --> 00:11:54.480 And then here we are. 00:11:54.799 --> 00:11:55.519 Congratulations. 00:11:55.759 --> 00:11:56.399 Thank you. 00:11:56.639 --> 00:12:02.000 Uh I am today leading all of the engineering team plus IT support and security. 00:12:03.440 --> 00:12:04.559 What does Z do? 00:12:05.120 --> 00:12:05.440 Okay. 00:12:05.679 --> 00:12:06.879 In simple terms. 00:12:09.919 --> 00:12:11.519 Okay, simple terms. 00:12:11.679 --> 00:12:24.000 Um
and Hada Shna Daimon and Modafin, Walla Numu Le Tujar. 00:12:24.559 --> 00:12:33.759 Uh Mahal offline, online, or this is a new distribution channel, that's it. 00:12:34.000 --> 00:12:36.720 That's nowhere near what we're thinking of in the Lamangul. 00:12:42.799 --> 00:12:50.879 Fashion, uh uh retail industry. 00:12:51.360 --> 00:12:54.720 They would benefit from different kinds of channels that they would be in. 00:12:54.879 --> 00:12:55.039 Okay. 00:12:56.720 --> 00:12:57.200 Fashion. 00:12:57.360 --> 00:12:59.039 And to fit online, bidata online. 00:12:59.200 --> 00:13:02.879 Like in your business really benefit from having a physical branch with pop-ups. 00:13:03.120 --> 00:13:04.720 We help you build POS. 00:13:04.960 --> 00:13:05.200 Okay. 00:13:05.600 --> 00:13:09.679 And then um so it's across the full uh scope. 00:13:10.000 --> 00:13:13.519 But that's why we introduce channels. 00:13:13.600 --> 00:13:19.120 So you can sell marketplaces, our own marketplace, Amazon, Trendiol, Jahez. 00:13:19.440 --> 00:13:23.679 Uh wherever we can allow our merchants to sell, we want them to sell. 00:13:24.559 --> 00:13:34.159 And in a guide, we say that okay, and your product is this kind, it's seasonal, it's uh uh niche, it's whatever, this these marketplaces that fit you. 00:13:34.240 --> 00:13:34.799 You don't have to be. 00:13:35.840 --> 00:13:36.720 You consult as well, yeah. 00:13:37.679 --> 00:13:38.080 Yes, yes. 00:13:38.240 --> 00:13:45.279 Uh but this is what made people uh closer to Z because we do consult and we do care. 00:13:45.519 --> 00:13:48.559 Uh we don't give them the technology and tell them okay, use it. 00:13:48.799 --> 00:13:51.360 Um we want them to actually grow. 00:13:51.600 --> 00:13:55.919 Um and it's a core value growth. 00:13:59.919 --> 00:14:15.039 Do you think retailers uh they push a decision based on uh bad habits or bad uh consequences and let's say a physical store in a physical store he doesn't have very good marketing. 00:14:15.440 --> 00:14:19.120 Obviously, he wants to expand, he wants a big exposure, let's say. 00:14:19.279 --> 00:14:22.399 He can't have it be a certain geographical location. 00:14:22.799 --> 00:14:24.720 Do you think Haida will directly push him? 00:14:24.879 --> 00:14:31.519 I'll be fakir akar negative label costs, and I saturate, I move, I expand. 00:14:31.919 --> 00:15:00.480 Obviously, it would be uh, tajern, online businesses um after offline uh online channel efficient fill operation how organization is not ready uh uh to be online. 00:15:01.200 --> 00:15:03.519 What the idea for retail is fulfillment. 00:15:03.679 --> 00:15:04.960 How do you fulfill orders? 00:15:05.120 --> 00:15:18.080 Uh if you're just gonna use uh as is and pick up uh items in the shelves and then that's not a um uh fee challenges. 00:15:18.320 --> 00:15:21.840 Um they have to streamline the process. 00:15:21.919 --> 00:15:25.440 This is why a lot of them do open online and then decide to stop. 00:15:25.600 --> 00:15:39.919 Uh they want the extra channel, but they don't want to do the effort to uh optimize the processes, and this is where the POS comes in, and uh uh POS keeps track of inventory in store and online and then allows you to pick up. 00:15:40.000 --> 00:15:47.440 Uh this is the any what we're trying to how do you how do you balance a product perfectly built? 00:15:47.919 --> 00:15:51.759 But market ready, and that's speed, perfection. 00:15:51.919 --> 00:15:58.320 How do you balance uh I think uh culture z we never aim for perfect? 00:15:58.480 --> 00:16:06.480 Uh we know that uh whatever we conceptualize will never match what actually is needed for market. 00:16:07.519 --> 00:16:21.279 But we iterate, we just build something that's good enough, iterate, hear back from our merchants, pilot a lot of the features we only launch with very small subset of our merchants, maybe 10 merchants, and then seek feedback and then improve and then launch. 00:16:21.840 --> 00:16:43.679 Some people might say this this makes Z products uh any uh less uh comprehensive and I would say it would make them on the short term less comprehensive because we we launch a small smaller feature, but on the long term they strike the pain point. 00:16:44.080 --> 00:16:45.759 Uh this is what I think this is. 00:16:46.159 --> 00:16:47.279 Okay, this makes sense. 00:16:53.360 --> 00:16:57.360 Fee concept technology would show our alam they are full. 00:16:57.440 --> 00:16:58.320 So we'll uptime. 00:16:58.720 --> 00:16:58.960 Okay. 00:16:59.919 --> 00:17:09.279 Uptime uh who up the minus system fee, must kill a fanal. 00:17:09.440 --> 00:17:11.599 This is any any in simple terms. 00:17:12.480 --> 00:17:17.680 How do you manage uh fee all 100% uptime? 00:17:17.839 --> 00:17:18.000 Okay. 00:17:20.079 --> 00:17:20.319 Yeah. 00:17:20.559 --> 00:17:22.400 Uh during peak times. 00:17:22.480 --> 00:17:25.759 Yeah, let's say uh Green Friday, Ramadan. 00:17:26.079 --> 00:17:27.039 How do you manage this? 00:17:27.200 --> 00:17:29.359 So is it is it actually achievable? 00:17:29.599 --> 00:17:36.960 Okay, so uh again, uptime is measured in time uh how much time of the year uh you're up in. 00:17:37.200 --> 00:17:41.519 Famia Belmiya means three uh 365 days you're always online. 00:17:42.720 --> 00:17:52.000 Uh the challenge is uh and most services are 99.9, a few nines after, depending on how much. 00:17:52.079 --> 00:18:01.440 But every nine means uh you're getting closer to less than a minute uptime, uh sorry, downtime, uh allowable downtime. 00:18:01.680 --> 00:18:09.839 And and every nine means had dunya mun can spell in terms of probability. 00:18:12.720 --> 00:18:21.599 Every nine they add sar, nine a little bad hazarli, nine albaad, it's yeah uh probability I get uh uh. 00:18:24.079 --> 00:18:27.759 We're talking about nine nine, ninety-nine point nine nine, not more than that. 00:18:27.920 --> 00:18:31.599 But this means we're only down for less than two minutes. 00:18:31.839
--> 00:18:36.319 Uh I think yeah, five or something in per per year. 00:18:36.559 --> 00:18:37.119 Wow. 00:18:37.279 --> 00:18:39.039 Uh this is what we're aiming for. 00:18:39.200 --> 00:18:41.119 Type how do we aim, how do we and yet? 00:18:41.839 --> 00:18:43.119 And alhamdulillah, we did it. 00:18:43.200 --> 00:18:49.359 Last year we did uh I think 99.98, and that was due to the Sarma Cloudflare. 00:18:49.759 --> 00:18:50.000 Okay. 00:18:50.319 --> 00:18:55.200 Uh we had nine uh seven months straight of 100% uptime. 00:18:55.519 --> 00:19:07.920 And this takes a lot of discipline, and I think this is what made Zed uh made Zed get to that stage we're at, is that we're disciplined with with how we manage things. 00:19:08.160 --> 00:19:13.920 Uh yeah, any long uh long gone the days are of yOLO uh of yoloing thing. 00:19:14.000 --> 00:19:14.640 Yellow khalas. 00:19:14.960 --> 00:19:16.559 Shag level production jarabish. 00:19:17.279 --> 00:19:24.480 Um we're more disciplined in how we manage deployments, we manage uh scaling, auto-scaling, and how manage rules. 00:19:24.640 --> 00:19:27.680 We know any seasons are seasons. 00:19:28.480 --> 00:19:30.880 They happen for yeah, we know when they happen. 00:19:32.480 --> 00:19:38.400 We know before Ramadan we should start having prepare ourselves, we pre-warm the environment and so on. 00:19:38.640 --> 00:19:41.519 Uh it's just discipline, one word. 00:19:43.279 --> 00:19:47.279 Fe complexity, obviously, to what you actually do behind the scenes. 00:19:47.839 --> 00:19:51.680 What's the complexity behind launching Z chip and ZPay? 00:19:52.319 --> 00:19:52.640 Okay. 00:19:53.680 --> 00:20:04.160 So Z chip was the first product that uh Z Khalinangul started outside of the Z core e-commerce thing. 00:20:04.400 --> 00:20:08.559 And I was the tech lead of that product back in 2020. 00:20:08.799 --> 00:20:14.480 Um, the challenge was um, first of all, this is not e-commerce. 00:20:14.559 --> 00:20:22.480 Uh well, it is e-commerce, but it's not uh eShop anymore, it's now logistics, it's not uh it's a different uh industry. 00:20:22.880 --> 00:20:31.920 But in terms of product uh thinking, uh we we all had to switch uh uh and shift our mindset. 00:20:32.240 --> 00:20:38.079 Uh and uh oddly, uh any of uh funny enough, I wrote about this in the uh Phil blog. 00:20:38.160 --> 00:20:41.119 We have a Z blog, Z engineering blog that we write about. 00:20:41.359 --> 00:20:46.880 But I was talking about how we serve the first uh 100,000 shipments with Z chip. 00:20:48.160 --> 00:20:51.680 Kind of challenge again is I need to understand logistics. 00:20:51.759 --> 00:20:59.440 It's no more uh exactly um how do carriers work? 00:20:59.519 --> 00:21:00.960 What are their APIs? 00:21:01.200 --> 00:21:03.200 How how do they assign orders? 00:21:03.359 --> 00:21:05.519 When do they assign orders to each carrier? 00:21:05.680 --> 00:21:11.440 Uh and then um what do I need to know about the customer so I can deliver something to them? 00:21:11.920 --> 00:21:18.880 International shipping, and then there was a lot of things that in terms of business acumen I needed to understand. 00:21:19.119 --> 00:21:23.920 And then in terms of engineering culture, was that we were building a new system. 00:21:24.000 --> 00:21:27.359 This is outside of the great monolith that Zid had. 00:21:27.920 --> 00:21:34.000 Um so how do we uh uh communicate between the two the two systems? 00:21:34.240 --> 00:21:38.160 And this what would this new system be hosted in the same environment? 00:21:38.240 --> 00:21:40.640 Well, hosted outside, what are the IPI boundary? 00:21:40.799 --> 00:21:51.920 What are the just these are the things that are I'm not taking able to uh to grasp how did you actually manage to get the new business scope or technical, etcetera? 00:21:52.240 --> 00:21:59.039 Yeah, uh there was a lot of let's try and see how this works. 00:21:59.200 --> 00:22:01.519 Uh There were uh principles. 00:22:01.599 --> 00:22:04.720 This is how uh I was any at least designing this. 00:22:04.960 --> 00:22:06.240 I'm setting these principles. 00:22:07.279 --> 00:22:11.759 Z uh should not do something anything uh specific to Z chip. 00:22:13.359 --> 00:22:17.359 Zid in that case was the original uh core system. 00:22:17.759 --> 00:22:22.240 Um so uh we were integrating other carriers. 00:22:22.480 --> 00:22:24.880 Z chip should be treated just like any other carrier. 00:22:25.279 --> 00:22:35.200 Okay, that makes sense, just to avoid the core team from getting more complexity to them, and then we focus on building that uh thing our ourselves and the zip chip team. 00:22:35.519 --> 00:22:54.799 And then in terms of z chip, their API has to be uh agnostic to carriers, and we want to build one unified interface so we can integrate carriers, and we got to a point where we had like 40 carriers integrated, and you were talking in uh RMX, SEMSA, and so on, uh the big names and the small names. 00:22:55.119 --> 00:22:55.440 Okay. 00:23:10.640 --> 00:23:22.480 We try to really be the centralized place where all of your orders can come in and like alhadafli. 00:23:23.599 --> 00:23:28.400 Okay, hella um rahdal bi hadith now, but 15 questions. 00:23:28.640 --> 00:23:28.799 Okay. 00:23:29.039 --> 00:23:30.240 How are 15 questions? 00:23:30.400 --> 00:23:32.240 Okay, they are rapid fire questions. 00:23:32.319 --> 00:23:35.519 So uh rasak no explanation. 00:23:35.759 --> 00:23:36.400 No explanation. 00:23:36.640 --> 00:23:37.680 We'll get back to some of them. 00:23:39.200 --> 00:23:40.400 But no explanation. 00:23:40.559 --> 00:23:43.680 If you have something to uh to mention, we'll mention it after it. 00:23:43.839 --> 00:23:44.079 Okay? 00:23:44.400 --> 00:23:44.559 Okay. 00:23:45.359 --> 00:23:50.160 So microservices or monolith. 00:23:50.400 --> 00:23:51.839 Monolith, all the way. 00:23:52.079 --> 00:23:54.559 Uh speed to market or code perfection? 00:23:54.799 --> 00:23:55.759 Speed to market.
00:23:56.319 --> 00:24:00.319 Which way Riyadh or Dubai for tech talent? 00:24:00.640 --> 00:24:02.079 Remote, everywhere. 00:24:02.400 --> 00:24:03.519 Riyadh or Dubai? 00:24:03.680 --> 00:24:04.160 Everywhere. 00:24:04.400 --> 00:24:04.640 Okay. 00:24:05.279 --> 00:24:05.599 Nice. 00:24:07.759 --> 00:24:08.400 Everywhere. 00:24:09.279 --> 00:24:12.880 AI and e-commerce, is it revolutionary or overhyped? 00:24:16.079 --> 00:24:19.200 At the moment, I think it's overhyped. 00:24:20.400 --> 00:24:20.960 At the moment. 00:24:21.279 --> 00:24:22.160 Keep it in your mind. 00:24:22.400 --> 00:24:24.960 Technical debt, is it a tool or a trap? 00:24:26.559 --> 00:24:27.200 Both. 00:24:32.240 --> 00:24:34.559 Remote work or engineering war room? 00:24:34.960 --> 00:24:35.759 Remote work. 00:24:36.720 --> 00:24:39.359 Data security or user convenience. 00:24:41.680 --> 00:24:42.720 Convenience. 00:24:46.160 --> 00:24:48.640 Global expansion or local dominance. 00:24:48.960 --> 00:24:50.000 Local dominance. 00:24:50.720 --> 00:24:53.440 The most used app on your phone, not called Zid. 00:24:53.759 --> 00:24:54.240 Slack. 00:24:54.559 --> 00:24:55.119 Slack. 00:24:55.920 --> 00:24:57.599 Saudi coffee or V60? 00:24:57.759 --> 00:24:58.480 V60. 00:24:59.039 --> 00:25:01.759 One word to describe Zid's engineering culture. 00:25:03.200 --> 00:25:04.160 Excellency. 00:25:04.720 --> 00:25:08.000 Is 2026 the year of fully automated retail? 00:25:09.440 --> 00:25:09.759 No. 00:25:11.440 --> 00:25:13.200 How can all 15 questions? 00:25:18.079 --> 00:25:18.400 Okay. 00:25:21.119 --> 00:25:23.440 Speed to market or code perfection. 00:25:23.920 --> 00:25:24.319 Yeah. 00:25:27.519 --> 00:25:29.200 I said speed to market. 00:25:31.839 --> 00:25:35.440 I think you'll never be perfect with your code. 00:25:36.480 --> 00:25:40.960 But you need to not put yourself in a position where you shot yourself in the foot. 00:25:42.799 --> 00:25:43.039 Yes. 00:25:43.920 --> 00:25:47.200 The goal is design something, and again, design. 00:25:47.920 --> 00:25:49.519 Some people just write code. 00:25:49.759 --> 00:25:57.759 I'm saying design things that are simple enough that it doesn't trap you in a very bad situation in the future. 00:25:58.240 --> 00:26:10.160 Best don't stress too much on making this so perfect because I never uh I do contrast writing with coding and writing good code. 00:26:16.960 --> 00:26:17.119 Okay. 00:26:18.880 --> 00:26:19.039 Okay. 00:26:19.279 --> 00:26:20.319 AI and e-commerce. 00:26:26.880 --> 00:26:28.720 AI is overhyped. 00:26:32.880 --> 00:26:33.200 Okay. 00:26:34.400 --> 00:26:41.279 I think ilal an uh model is not mature enough. 00:26:42.960 --> 00:26:44.799 I'm not talking about the AI as technology. 00:26:48.160 --> 00:27:02.480 Shopify and uh OpenAI, they worked on uh I forgot the name of the standard, but to allow the protocol uh the protocol that allows any feeds, product feeds to show up on we'll add the name to it. 00:27:02.640 --> 00:27:03.440 We won't worry about it. 00:27:03.759 --> 00:27:10.640 Uh on on uh on uh chat bots or any chat uh uh models. 00:27:11.039 --> 00:27:13.279 Uh yes, it it's there. 00:27:13.519 --> 00:27:24.640 But use case mahiwab habi any it's not uh it's not mature enough so that we say in uh in this year would be the year that everyone shops through uh LLMs. 00:27:24.880 --> 00:27:36.160 Um and then the checkout model, I think recently also Shopify added four percent to uh as a fee on the merchant for every purchase that happens through open AI. 00:27:36.880 --> 00:27:38.160 Just a lot of too much. 00:27:38.480 --> 00:27:38.880 4%. 00:27:39.759 --> 00:27:44.480 Uh again, uh is expensive. 00:27:44.640 --> 00:27:51.519 Uh and uh that now that it doesn't really really make sense for merchants to show up on the models. 00:27:51.599 --> 00:27:53.599 But I think this is why it's overhyped. 00:27:53.680 --> 00:28:02.319 It's not that the technology is overhyped, though I think parts of the technology are overhyped, but in general, I think we're not yet there in terms of a business model. 00:28:02.640 --> 00:28:02.960 Okay. 00:28:03.599 --> 00:28:06.799 Um data security or user convenience. 00:28:07.119 --> 00:28:15.359 There is an asterisk if we there are minimum minimum security requirements that should always be met. 00:28:15.599 --> 00:28:31.599 So you're willing to move on with the minimum, which makes the app as the actual experience safe, as long as you you you can tick the boxes that says uh my customer is happy, the behavior of my customer is valid, etc. 00:28:31.759 --> 00:28:32.000 etc. 00:28:32.319 --> 00:28:35.200 Yeah, I'm not saying be sloppy. 00:28:35.279 --> 00:28:40.720 You should really consider cybersecurity any very seriously, and this is what we do at Zit. 00:28:40.880 --> 00:28:44.960 But there are the sort of any steps that are unnecessary. 00:28:45.200 --> 00:28:49.200 Say, if you want to change users' permission, send an OTP. 00:28:52.720 --> 00:28:55.440 Too much OTPs, yeah, too much OTPs. 00:28:56.160 --> 00:29:05.920 Uh make sure that the whatever you're doing is not catastrophic in terms of cybersecurity and focus on convenience. 00:29:06.160 --> 00:29:11.759 Um authenticate a session as long as you've already OTP'd for this session, keep the whole session on. 00:29:12.400 --> 00:29:14.160 Uh these are the sort of things that I meant. 00:29:14.400 --> 00:29:15.200 Not being sloppy. 00:29:15.279 --> 00:29:17.200 Sloppy is bad in general. 00:29:18.160 --> 00:29:19.039 You don't want to be there. 00:29:19.200 --> 00:29:23.920 But focus on how do you make integrate security so that it's seamless. 00:29:24.160 --> 00:29:27.359 You know you know something, and um mentari it had hakiak. 00:29:28.640 --> 00:29:42.240 It looks like you have full trust, capacity, leader team, anything that might pop up, we're we're able to actually fix it on spot, time frame. 00:29:43.119 --> 00:29:43.279 Yeah. 00:29:47.359 --> 00:29:48.400 How is it?
00:29:48.720 --> 00:29:51.680 So I am all for empowerment. 00:29:51.920 --> 00:30:00.079 Uh I I don't like me personally being the bottleneck, nor any one of the leads that report to me being the bottleneck. 00:30:00.319 --> 00:30:08.559 Uh I start by my style with the leads that report to me, that then hopefully then trickle down to the rest of the team. 00:30:08.799 --> 00:30:25.119 But what I do is in uh Metharan, if you're leading uh this um uh part of the team, say product engineering, you're the head of product engineering, you're in you build you basically this is the team that builds uh zit. 00:30:25.519 --> 00:30:33.519 Um I am not gonna be on top of your head every day asking what the team did today and are they doing things well enough. 00:30:34.640 --> 00:30:35.759 I'm not micromanaging. 00:30:36.000 --> 00:30:43.200 But micromanagement doesn't mean uh and this is where you should balance. 00:30:43.279 --> 00:30:46.319 Um, there are rules that I'm gonna keep track of things. 00:30:46.480 --> 00:30:58.160 I'm gonna ask you because you're responsible uh and you have ownership, but you're empowered to do things that you without my without like referring back to me or taking my permission. 00:30:58.400 --> 00:31:04.559 But usually too often I have a one-to-one with that uh head of department every two weeks, not every week. 00:31:04.799 --> 00:31:11.119 Uh, because I want to give them enough time to make decisions, maybe make wrong decisions, and then come back and report back. 00:31:11.519 --> 00:31:12.319 Wrong is healthy. 00:31:12.480 --> 00:31:15.279 Yes, wrong is healthy because this is how I learned. 00:31:15.359 --> 00:31:25.519 And uh I you can for a great uh majority of my career, I didn't have a um uh a direct manager that was so technical. 00:31:25.759 --> 00:31:32.480 Um uh or or even at some point we I didn't have a direct manager that was engineering at all. 00:31:32.720 --> 00:31:35.839 But I was making all the decisions myself. 00:31:36.000 --> 00:31:43.119 Uh and this is how I learned and then why not? 00:31:43.359 --> 00:31:43.839 Makes sense. 00:31:43.920 --> 00:31:48.160 Yeah, how do you retain top talent Sustainable market? 00:31:48.480 --> 00:31:53.519 It's a very competitive market, fintechs, uh huge uh corporations. 00:31:53.680 --> 00:31:55.359 How do you retain your team? 00:31:57.119 --> 00:31:58.880 Uh it's a difficult question. 00:31:59.119 --> 00:32:04.000 Uh so there are a few principles that I set uh for myself. 00:32:04.240 --> 00:32:08.160 I think money is one key. 00:32:08.400 --> 00:32:09.839 Let's get this out of the way. 00:32:10.000 --> 00:32:18.960 Yes, the biggest key, but assuming you're good with the uh with the pay, uh, there are other things that keep people comfortable working at this company. 00:32:19.119 --> 00:32:25.279 Uh, one of them being uh uh semiha technical maturity and engineering culture. 00:32:26.160 --> 00:32:28.960 Uh we have some zidders who left Zid. 00:32:29.440 --> 00:32:30.240 Hello, zidders. 00:32:30.480 --> 00:32:31.440 Yeah, we call them zidders. 00:32:31.680 --> 00:32:31.920 Nice. 00:32:32.000 --> 00:32:34.400 Uh we had some zidders who left Zidd. 00:32:34.559 --> 00:32:41.759 Uh the shortest uh departure time was a month, and then they came back. 00:32:42.000 --> 00:32:42.319 Okay. 00:32:42.960 --> 00:32:46.960 Uh that person went to a European company. 00:32:47.200 --> 00:33:00.319 He assumed in uh it's in Europe, it's gonna be a very mature uh technical, uh, technically any, and then they came back to Z, uh any cursing at them, and they were like, I don't know how how they're operating. 00:33:00.799 --> 00:33:04.400 They're bad code linting, they don't even have code linking. 00:33:04.480 --> 00:33:19.680 When I installed code linting, uh code linter to them, they were like, Leish and uh kidah, this breaks my code, this uh uh throws an error, uh, throws errors, or they they didn't even know how to run a product linter. 00:33:19.920 --> 00:33:21.119 Uh sorry, the code linter. 00:33:21.839 --> 00:33:30.000 Uh and the deployment process was bad, and they were like, any, the this company is nowhere near what I was doing, that's it. 00:33:30.400 --> 00:33:32.799 And the the challenges are not so interesting. 00:33:33.039 --> 00:33:39.279 Um for talented people and high caliber, they want to be challenged. 00:33:39.519 --> 00:33:44.960 Um it's very rare that I see someone who's like, I want a laid up a laid-back job. 00:33:45.279 --> 00:33:46.480 But we have all the challenges. 00:33:48.240 --> 00:33:52.319 The common example that I give to people at Zidden uh we cannot go down. 00:33:52.480 --> 00:33:55.759 We cannot go down for a maintenance, not go down as in outage. 00:33:56.079 --> 00:33:57.200 We cannot go down for maintenance. 00:33:57.359 --> 00:33:58.640 How do you do your maintenance? 00:33:58.880 --> 00:34:01.839 Uh we do maintenance without maintenance windows. 00:34:02.799 --> 00:34:06.960 We all the maintenance we do is up uh uh we call what we call online uh maintenance. 00:34:07.920 --> 00:34:15.280 You keep the system uh the system up and you do things that do not block uh block the the uptime, and this is where discipline comes in. 00:34:15.679 --> 00:34:25.360 Uh these need to be inter handled intricately so that it doesn't cause an outage. 00:34:25.599 --> 00:34:28.480 Uh this is a challenge. 00:34:28.559 --> 00:34:32.239 Uh not every company has, which makes it fun for me. 00:34:33.360 --> 00:34:36.880 Uh I can say no, I cannot go down. 00:34:37.360 --> 00:34:37.920 It's very nice. 00:34:38.159 --> 00:34:47.920 No, you do maintenance, you do online maintenance, you don't go down, you have an uptime of 99.98 0.01 when uh cloud fit. 00:34:48.239 --> 00:34:48.639 Yes. 00:34:48.880 --> 00:34:52.239 Okay, uh the government is pushing digital payments. 00:34:56.639 --> 00:34:57.039 Yes. 00:34:57.519 --> 00:35:03.679 Um it was helpful, push any COVID was the big uh push. 00:35:04.000 -->
00:35:08.719 Uh pre-COVID, um orders, cash and delivery were very high. 00:35:08.880 --> 00:35:11.280 Sarah and uh was a surprising number. 00:35:11.440 --> 00:35:19.199 Um I I considered myself that I any I'm early adopter of just not having a wallet at all. 00:35:19.360 --> 00:35:24.239 I do have a wallet, but saying I don't have cash on the in the wallet, but not having cash at all. 00:35:24.800 --> 00:35:29.760 And for some reason I assumed my circle is representative of the rest of Saudi. 00:35:30.320 --> 00:35:32.000 Uh and then I saw the numbers. 00:35:36.639 --> 00:35:41.440 I saw the numbers were like 60% or 50% at the time orders were cash and delivery. 00:35:43.119 --> 00:35:51.920 That was a big segment of the market, and this was one of the influencing principles for Z chip, and then we wanted to handle cash and delivery collection for merchants. 00:35:52.239 --> 00:35:55.840 Um that was surprising. 00:35:56.000 --> 00:36:11.840 Uh, and as much as they wanted to push online payment and make it more streamlined and use Apple Pay bus, still merchants were not really any uh sorry, not merchants, customers, and customers were not really trusting some random website on the internet taking my money. 00:36:12.320 --> 00:36:15.519 What's the percentage now of people actually using cash on delivery? 00:36:16.480 --> 00:36:18.719 To be honest, I don't have an updated number. 00:36:18.880 --> 00:36:23.519 Uh I stopped tracking that once we we got to a sufficient anyway approximately. 00:36:24.159 --> 00:36:25.360 I think that's less than 10%. 00:36:26.000 --> 00:36:26.800 Wow, yeah. 00:36:27.039 --> 00:36:33.440 Um again, COVID, then push for Made Pay, uh being really out there. 00:36:33.679 --> 00:36:54.159 Um now with Samsung Pay, with uh with Google Pay, halas any there is no reason for the merchant uh customers not to use uh uh online payments, and now with Tamara and Tabby and other uh buy now pay later services, halas uh people do prefer uh buy uh using digital payments than cash. 00:36:54.719 --> 00:36:56.159 How are you aligning hello? 00:36:57.920 --> 00:37:00.320 How are you aligning Malvision 2030? 00:37:00.480 --> 00:37:02.800 Yeah, it's uh I think it's something massive. 00:37:03.519 --> 00:37:07.679 We're aligning very well, but this way you take headlines, but in depth. 00:37:11.519 --> 00:37:12.159 Headlines. 00:37:12.400 --> 00:37:18.320 I I mean the general principle we're increasing the GMV. 00:37:18.559 --> 00:37:34.800 Uh GMV uh of Saudi, it's not that us doing it, but we're playing a big role or an important role as being a platform that pushes merchants to grow and allow them to do more transactions. 00:37:35.440 --> 00:37:50.400 I think that's one of the places that we're playing a role in, and obviously being a tech company in Saudi localizing technology and uh and a lot of what we do, design and build is done here in Saudi, so uh that also localizes technology in Saudi. 00:37:50.719 --> 00:37:51.039 Okay. 00:37:51.679
--> 00:37:54.719 Stamlo Google Cloud briefly. 00:37:55.280 --> 00:38:11.519 On um so because we're trying to localize uh up sorry for data residency, we're trying to uh provide uh local hosting for our customers because uh we started with AWS and then uh we wanted to provide a local hosting for our customers. 00:38:11.840 --> 00:38:14.000 So Google they do have the data center built up. 00:38:14.320 --> 00:38:22.880 Okay, so we we we allow our merchants to choose, they have to apply for it, but say and they they can choose a region to be in Saudi through Google Cloud. 00:38:23.119 --> 00:38:40.719 Complexity, VP of engineering, uptime, uh 99.98, uh online maintenance, but they are for uh how can we wrap our brain a la fiqhirit and no into my kilatah, my team, my new initiative. 00:38:42.000 --> 00:38:48.800 How do you unplug knowing and no fiand customers relying on you 247 365? 00:38:49.360 --> 00:38:51.440 Uh it's funny. 00:38:51.599 --> 00:38:53.920 Uh I have three hobbies. 00:38:54.159 --> 00:38:56.320 Three very different hobbies. 00:38:56.639 --> 00:39:01.599 Uh one is uh reading. 00:39:01.840 --> 00:39:04.719 Uh I do read philosophy. 00:39:06.000 --> 00:39:16.800 These are the and usually especially if you go very theoretical, you can go things that are very detached from the real world. 00:39:17.519 --> 00:39:20.480 This gives me an opportunity this to disconnect. 00:39:20.800 --> 00:39:32.000 Uh and then if I if I'm not feeling actually Aisha Alakab Harikulmuch, uh I watch uh movies, Egyptian movies. 00:39:32.320 --> 00:39:36.880 Uh and then what's your favorite Egyptian movie? 00:39:37.760 --> 00:39:41.280 Uh that's a difficult question. 00:39:41.440 --> 00:39:43.280 Uh I don't re-watch movies. 00:39:43.599 --> 00:39:43.840 Okay. 00:39:44.719 --> 00:39:53.199 Usually what's uh what I like about Egyptian movies is that they're not necessarily about the whole the storyline ever. 00:39:53.440 --> 00:40:02.719 They're about if you just listen to the uh scenario, you'll laugh, that's it. 00:40:02.960 --> 00:40:03.679 Good enough. 00:40:03.920 --> 00:40:08.159 Uh rewatching movies isn't as funny. 00:40:08.639 --> 00:40:10.239 That's why I don't re-watch movies. 00:40:10.719 --> 00:40:11.679 Tell it hobby. 00:40:12.000 --> 00:40:14.880 Uh it's an expensive hobby. 00:40:16.639 --> 00:40:17.760 I like classic cars. 00:40:18.079 --> 00:40:19.280 Do you collect classic cars? 00:40:19.440 --> 00:40:19.760 I don't. 00:40:20.079 --> 00:40:20.960 It's expensive hobby. 00:40:21.119 --> 00:40:21.440 Not yet. 00:40:21.599 --> 00:40:23.360 But hopefully at some point. 00:40:23.519 --> 00:40:26.559 But I like to like go and see uh classic cars. 00:40:26.880 --> 00:40:28.480 What's your dream classic car? 00:40:29.039 --> 00:40:29.599 Targa. 00:40:30.079 --> 00:40:31.679 Some 80s targa. 00:40:31.760 --> 00:40:33.760 Not yet uh sure which one. 00:40:34.079 --> 00:40:35.039 Nice, nice. 00:40:35.280 --> 00:40:41.920 But the one technical milestone and tahazan send me a bucket list, I'll checklist. 00:40:42.239 --> 00:40:45.199 You want to tick by the end of 2026? 00:40:46.480 --> 00:40:51.440 So uh in the past two years, we rewrote a lot of our systems. 00:40:51.599 --> 00:40:56.880 Uh again, moving and we want to build something that would take us for the next five years. 00:40:57.119 --> 00:41:01.760 Uh given that a lot of these systems were built eight years ago. 00:41:01.920 --> 00:41:08.320 Uh they're especially my MVPs and so on, they're they're not serving us. 00:41:08.480 --> 00:41:20.960 Uh we rewrote two very big systems, storefront system that renders the themes and allows the merch uh the customers to actually buy from Z and uh uh and interact with Zed. 00:41:21.199 --> 00:41:23.199 And we rewrote our merchant dashboard. 00:41:23.360 --> 00:41:27.039 This is what our merchants use to manage their stores and orders. 00:41:28.079 --> 00:41:35.440 And now we have 1.2 million lines of code that I'm planning to purge sometime this year. 00:41:35.760 --> 00:41:41.679 Uh if I take that box, uh, I'm gonna make a very big party. 00:41:41.920 --> 00:41:42.480 That's it. 00:41:42.719 --> 00:41:43.599 We're gonna be there. 00:41:43.840 --> 00:41:49.119 Obviously, and yeah, I'm gonna do something outdoors in the in front of this uh the office. 00:41:49.360 --> 00:41:50.960 I'm gonna do something very huge. 00:41:51.119 --> 00:41:53.360 So you everyone is welcome to uh join us. 00:41:54.480 --> 00:41:55.599 Nice, nice. 00:41:55.760 --> 00:42:00.000 Uh uh it was Anjad, uh a huge pleasure. 00:42:00.320 --> 00:42:03.599 Insights, you you have amazing insights. 00:42:03.760 --> 00:42:06.719 You're hke simple, uh, fun. 00:42:07.599 --> 00:42:08.400 I love it. 00:42:08.559 --> 00:42:09.599 I love the session. 00:42:09.760 --> 00:42:16.960 Um, I hope um in that was lovely, Sarah, and I enjoyed it. 00:42:17.199 --> 00:42:17.599 Perfect. 00:42:17.760 --> 00:42:19.360 Thank you, Anukatman al Yom. 00:42:19.679 --> 00:42:20.559 This was unplugged. 00:42:20.960 --> 00:42:22.320 Thank you, absolutely.
Editar transcrição