Engaging with Technology Expert, Dr. Edwin Hernandez

WEBVTT
00:00:03.225 --> 00:00:07.732
This episode is brought to you by Roundtable Group the experts on experts.
00:00:07.732 --> 00:00:11.685
We've been connecting attorneys with experts for over 25 years.
00:00:11.685 --> 00:00:15.432
Find out more at roundtablegroupcom.
00:00:18.221 --> 00:00:19.969
Welcome to Discussions at the Roundtable.
00:00:19.969 --> 00:00:26.001
I'm your host, noah Ball, and today I'm excited to welcome Dr Edwin Hernandez to the show Now.
00:00:26.001 --> 00:00:37.276
Dr Hernandez is the founder of EGLA Corp, a full-service IP and tech consulting and investment firm, and EGLA Vader, a tech incubator and accelerator.
00:00:37.276 --> 00:00:45.134
He's an expert in AI and machine learning, wireless software technologies and serves on multiple IEEE boards.
00:00:45.134 --> 00:00:50.627
Dr Hernandez holds a PhD in computer engineering from the University of Florida.
00:00:50.627 --> 00:00:52.430
Let's jump into it.
00:00:52.430 --> 00:00:56.633
You've had an impactful tech career with a wide range of accomplishments.
00:00:56.633 --> 00:00:59.741
How did you first get started as an expert witness?
00:01:01.543 --> 00:01:04.007
I began in 2014.
00:01:04.007 --> 00:01:09.956
A friend of mine is a professor at the University of Florida and he was unable to take care of one case.
00:01:09.956 --> 00:01:14.284
So he says, like recommended me to take care of the case.
00:01:14.284 --> 00:01:20.411
I successfully was able to help this customer in a trade secret manner and that's how everything began.
00:01:20.411 --> 00:01:22.013
And that's how everything began.
00:01:23.313 --> 00:01:25.837
Did you get any mentorship from your friend?
00:01:25.837 --> 00:01:32.832
Did you have some idea of what to expect for your first couple of times as an expert witness, or were you kind of just thrown into it?
00:01:34.100 --> 00:01:46.939
No, I actually had a prior experience where I was kind of like an expert not precisely on the full flesh, it was my own patent case, honestly.
00:01:46.939 --> 00:01:48.063
Uh, on the full flesh, uh, it was my own patent case.
00:01:48.063 --> 00:02:06.382
And then in that case I learned the tricks of the, of the, of the game, basically like all the, the filings, the motions and have everything worked in um, in federal circuit uh, anything that you'd care to share, any of those tricks of the trade that might be helpful to others, particularly newer expert witnesses who are just getting started.
00:02:07.945 --> 00:02:14.586
Well, I will say that the main thing for an expert is to really stick with something very fundamental the truth.
00:02:14.586 --> 00:02:38.546
So if you stay truthful, you stay with the truth, you stay with the facts and you say something factually based, that you have some sort of a basis either that could be corroborated by some sort of your knowledge, your experience, if you have a vast amount of experience in certain fields.
00:02:38.546 --> 00:02:53.532
Even in those cases, you need to start pointing out in the right direction to factual data that you find from the discovery, which discovery means that you're identifying all the facts of the case.
00:02:53.532 --> 00:02:59.193
So I think that's I would say that's the main thing to stay valid, let's say, in this business.
00:02:59.193 --> 00:03:06.913
Otherwise, quickly it will be that word motion and most likely have some issues there.
00:03:08.281 --> 00:03:09.667
Let's talk about those calls.
00:03:09.667 --> 00:03:12.909
So somebody calls you on the phone, says I'm looking for an expert witness.
00:03:12.909 --> 00:03:20.271
What are the sorts of questions that they ask you and what are the sorts of questions that you ask them to determine if it's going to be a good engagement or not?
00:03:21.700 --> 00:03:46.468
Well, the main thing that I ask or I get asked for is, uh, my background, what have I done in the past, what kind of cases I've been working on and um, and basically the challenges that I had in in different circuit, cases where I had um motions to strike, that were hearings and things like that, and how I survived those.
00:03:46.468 --> 00:03:51.929
Basically, that has been mostly one, because I have too many cases in my, in my pocket.
00:03:51.929 --> 00:04:18.413
So then uh, and I will say that that's uh, one of the things I get asked a lot and I ask always the question is it was a time trip because, as you know, sometimes they call you like oh, you need to report like a month and you need to review like thousands of pages of something, and I cannot draft like an opinion on something in a very short period of time because I'm not full time using that case.
00:04:18.413 --> 00:04:27.699
Right, even if I was, in that case, full time, it would require ample of time to get used to it case full time, it will require ample of time to get used to it.
00:04:28.660 --> 00:04:34.810
That's something that I've heard from a lot of my interviewees that they get a lot of calls for kind of rush jobs Like I need you to read 6 trillion pages and I need a report on it tomorrow.
00:04:35.521 --> 00:04:39.386
You need to request an extension of time and sometimes they do.
00:04:39.386 --> 00:04:57.144
Sometimes I'm in a rush to the main thing that you have to calculate, calculate how many patents that you have to review, how many pages of code, software, technical specifications are behind each of those patents, and then can that be done at a time frame requesting to be done?
00:04:57.144 --> 00:05:13.644
Because many times I wish we have an AI tool that could just see that and say, look, don't do it, but but those four, I think tool that could just see that and say like, look, don't do it, but by thus far I think it's still like an expert.
00:05:13.644 --> 00:05:14.745
Estimates on time is because you need to commit to.
00:05:14.745 --> 00:05:15.668
I mean, people don't understand.
00:05:15.668 --> 00:05:23.004
It's all right in the report, it's surviving the summary judgment and surviving the entire depot, right, that's the main thing.
00:05:24.209 --> 00:05:28.204
Sure, you know you mentioned and obviously you're an expert in AI.
00:05:28.204 --> 00:05:30.149
Where are we going with AI?
00:05:30.149 --> 00:05:32.136
What does the future hold in AI?
00:05:32.136 --> 00:05:36.744
Is it a passing trend or is it something that's here to stay?
00:05:36.744 --> 00:05:39.005
Has Pandora's box already been open?
00:05:40.206 --> 00:06:13.872
Well, to be honest with you, I joined the AI revolution in 1996 to 2000, when I was part of the machine intelligence lab at the University of Florida and I began learning everything on machine learning, then neural networks and LISP and all kinds of different paradigms that existed at the time and Lisp and all kinds of different paradigms that existed at the time and for so many years I actually never surfaced that expertise because it looked like it was a joke.
00:06:13.872 --> 00:06:16.949
You look less serious.
00:06:16.949 --> 00:06:18.786
You say I'm a robotics expert.
00:06:18.786 --> 00:06:23.706
I'm an AI expert 2000 to 2010,.
00:06:23.788 --> 00:06:34.990
That was most likely a recipe for a disaster interview or in any job or consulting job you would not be taken seriously because AI was not really working.
00:06:34.990 --> 00:06:48.872
And recently, with the GPT changes and if I studied GPT substantially, I believe that the changes made are not fundamental to a complex artificial intelligence paradigm.
00:06:48.872 --> 00:07:19.889
It's just basically very smart ways to twist and turn the way you use data, like some sort of encoding to review the data that's coming to open new ways and new language models, a lot of language models that are being accumulated to the future.
00:07:19.889 --> 00:07:41.766
So I think what they opened the Pandora box well, this is not necessarily like I think is AI percent, because you can make AI very complex is the way those engineers creatively connected AI with basic functionality that existed in the past in coding and some other practices that connected the two dots.
00:07:41.766 --> 00:07:45.206
And then now we have this new revolution with GPT models.
00:07:46.442 --> 00:07:48.891
And it's quite a broad topic.
00:07:48.891 --> 00:07:53.220
How do you remain an expert in your really multiple fields?
00:07:53.220 --> 00:08:02.130
What do you do to stay on top of everything, with all of this cutting edge technology that's changing, sometimes is very rapidly, sometimes on a daily basis?
00:08:02.130 --> 00:08:05.425
What does it mean to be an expert in your field?
00:08:08.670 --> 00:08:31.504
Well, I mean, I think I would recommend, if you want to remain an expert number one is to attend conferences, technical conferences from IEEE, join the technical committees in different conferences, review papers, recent papers from people that will keep you updated shows, mobile World Congress, natural Association, broadcasting Show.
00:08:32.044 --> 00:08:50.639
So then you attend those conferences and you see what's going on at the time technically, the state of the art, and, in my case, I run an incubator, so we incubate companies, so then we invest in different startups, so I see a plethora of like maybe.
00:08:50.639 --> 00:09:05.922
Well, in december has been kind of sluggish, I've been doing all those things, but let's say, in this, this, this 2023, I reviewed maybe like around 80, 80 to 100 business plans 80 to 100, yes, a spot with multiple people in different events.
00:09:05.922 --> 00:09:35.604
So then you get up to speed to see what people are going through with the technology, where I mean you, you in five, in in 30 seconds, you know the idea is is novel or not, and then, obviously, like being in all these patent cases is uh, keeps you up to speed, keeps you updated, because you get to see a source code, you get to review the uh schematics, you get get to review multiple things that make yourself be current and cutting edge.
00:09:37.870 --> 00:09:41.682
Let's jump back into some of the specifics of your expert witnessing work.
00:09:41.682 --> 00:09:45.139
Have you worked for both the plaintiff and defendant side?
00:09:45.139 --> 00:09:47.154
I have.
00:09:47.235 --> 00:09:50.620
Actually I have worked for defendant and plaintiff.
00:09:50.620 --> 00:09:55.480
In the plaintiff side I think is being a lot of patent cases in.
00:09:55.480 --> 00:10:05.721
On the defendant side I had some three secret background on a trade secret matter and major was to demonstrate that the evidence shown was not a trade secret.
00:10:05.721 --> 00:10:24.960
Obviously I cannot discuss the details because it was under a protective order and I was really surprised, very interestingly, because code reviews usually take a long time and you have to travel to a law firm and be what is called in a clean room environment where you don't have your laptop, you don't have your phone.
00:10:25.389 --> 00:10:26.716
They had a very interesting tool.
00:10:26.716 --> 00:10:28.735
They had a biometric tool that you will.
00:10:28.735 --> 00:10:32.235
They would run your biometrics and then you will see.
00:10:32.235 --> 00:10:45.592
On a PC actually, you will connect to a remote site and then if you will turn your head to the right or to the left for way too long, then it will log out and then it will track how many minutes you were reviewing the code.
00:10:45.592 --> 00:10:53.562
Well, I mean, the tool didn't allow me to move the head to the right or to the left, so I don't think either the tool was wrong or I reviewed 17 hours of code, right.
00:10:53.950 --> 00:11:17.599
But obviously, like I said, you have to be very concentrated, very like you need to concentrate your efforts in your mind to really really help the customer review that data that is being presented to you in a very thorough way, because, like I said, the deposition time is when you have to have the most amount of knowledge to defend your position.
00:11:17.599 --> 00:11:20.720
Because depositions is not, you don't win anything.
00:11:20.720 --> 00:11:22.576
I mean you're trying not to lose.
00:11:22.576 --> 00:11:24.918
That's kind of like the game in depositions.
00:11:24.918 --> 00:11:50.741
So I think that's very interesting how I came about helping both plaintiffs and defendants in cases in trademark, trade secret cases, patent cases and copyright violations in some cases where technology is involved in most of those situations, have you found yourself working on teams with other expert witnesses?
00:11:50.870 --> 00:11:52.135
What are those interactions like?
00:11:53.532 --> 00:11:54.072
yes, I did.
00:11:54.072 --> 00:11:58.881
Actually, the last case I was uh with uh, two more experts.
00:11:58.881 --> 00:12:04.923
Usually you have to have a damages expert and I am a technical expert, so you have a team right there.
00:12:04.923 --> 00:12:21.128
Uh, the damages expert rely on my expertise to do some of their calculations, for example, and in some of the cases I am a part of a team of two or three experts where we review multiple patents.
00:12:21.128 --> 00:12:33.644
They assign three or four to me, three or four to the other guy and then whatever, and then we both have to collaborate because in some cases the patents are interconnected.
00:12:33.644 --> 00:12:38.058
They have similar claims, they have similar technologies.
00:12:38.058 --> 00:12:39.775
We're reviewing the same evidence.
00:12:39.775 --> 00:12:54.647
So, obviously, whatever I propose in some cases as part of solving one of the claims needs to be involved with the other experts as well to really get to an agreement.
00:12:54.647 --> 00:13:03.706
Yeah, that's the path to go, or there's too much risk because of whatever arguments are being made and invalidity as well.
00:13:03.706 --> 00:13:08.322
As you know, it's a big gamble.
00:13:08.322 --> 00:13:12.941
Right, validity and infringement usually are very tied together.
00:13:14.169 --> 00:13:14.871
Right, of course.
00:13:14.871 --> 00:13:18.793
And when you're working with another team of experts, you're talking about collaboration.
00:13:18.793 --> 00:13:21.058
Are you also collaborating with the attorney?
00:13:21.058 --> 00:13:23.673
Do you feel like you're all kind of a team?
00:13:25.238 --> 00:13:28.004
Of course attorneys are great.
00:13:28.004 --> 00:13:32.495
I mean, I usually would like the attorney to tell me more of the legal side of the story.
00:13:32.495 --> 00:13:38.215
What arguments are being said by the other side, non-infringement arguments, for example.
00:13:38.215 --> 00:13:41.202
Sometimes the non-infringement arguments are ridiculous.
00:13:41.202 --> 00:13:45.181
They're not even possible to be technically done.
00:13:45.181 --> 00:13:56.202
So my job is to help the attorney understand what's within the boundaries of that and obviously then assist counsel with discovery as well.
00:13:56.830 --> 00:13:59.210
Sometimes discovery could become really problematic.
00:13:59.210 --> 00:14:17.821
I have one case that I discovered that they were hiding information and then attorneys were able to get the information, even a few years later able to get the information even a few years later and then obviously those attorneys will get sanctioned because of my discoveries.
00:14:17.821 --> 00:14:35.254
And that was right, because the suspicion that I had about the hiding of the information was true and the evidence showed in a supplemental report that there was a lot of evidence that was missing from the production initially, report that there was a lot of evidence that was missing from the production initially.
00:14:35.254 --> 00:14:47.224
And those attorneys and that client the opposite client I don't know if they received sanctions, but there was a motion for sanctions because of that failed to produce important documents.
00:14:48.404 --> 00:14:56.090
Wow, that's an amazing story no-transcript.
00:14:56.095 --> 00:14:58.756
Yeah, I have one right now actually.
00:14:58.756 --> 00:15:19.688
And then yesterday actually we received had a double hearing for that particular case and in fact they were questioning my expertise to determine some percentages and some of the facts for the case that they were being challenged, but the judge denied their motion completely.
00:15:19.688 --> 00:15:21.370
So now I get.
00:15:21.370 --> 00:15:37.539
Well, I don't have the I won't call it the opportunity, but I will have the right to tell the jury in trial at the end of January, the story in full, without any impediments.
00:15:37.539 --> 00:15:38.916
How do you?
00:15:38.975 --> 00:15:48.967
deal with those sorts of situations where somebody is trying to impeach your credibility or your ability to opine on a specific topic?
00:15:48.967 --> 00:15:56.369
Do you have any strategies that you use to deal with being impeached either in a deposition or in front of a jury?
00:15:57.495 --> 00:15:59.480
Yes, of course I mean that happens a lot.
00:15:59.480 --> 00:16:07.883
Attorneys will try to impeach you try to strike certain things that you cannot say, or you can say only certain things.
00:16:07.883 --> 00:16:16.975
But I think the main thing you have to go back to the beginning, go back to the discovery, go back to that report the day you drafted the report.
00:16:16.975 --> 00:16:22.206
Do you have sufficient amount of back back information?
00:16:22.206 --> 00:16:26.000
You have any sufficient amount of facts that will?
00:16:26.000 --> 00:16:35.481
Will that paragraph that says I opine that this is true or this is false or this is, uh, this is infringing on certain things, and that okay?
00:16:36.263 --> 00:16:56.104
Okay, that's because paragraph n minus one through n minus seven, whatever it was, then have sufficient amount of of content, omit that you can rely on, because once they go that day to impeach that paragraph and you will be no sir like you, you have a minus two.
00:16:56.104 --> 00:16:58.188
This causes part of that in minus four.
00:16:58.188 --> 00:17:14.259
So then you will see the and then with those power, we have links to exhibits, we have links to bait numbers, we have links to things that I have reviewed and and that's the strategy to get to not getting pitch doesn't happen, post happens free.
00:17:14.259 --> 00:17:23.201
So when you're writing the report, you're preparing for that, because once the report is written and once you have supplement, everything has that's been filed.
00:17:23.201 --> 00:17:24.144
Discovery is closed.
00:17:24.144 --> 00:17:25.968
There's no way to correct that.
00:17:25.968 --> 00:17:29.305
You can clarify that, but you cannot add any more arguments.
00:17:30.615 --> 00:17:33.125
Let's talk about your report writing process a little bit.
00:17:33.125 --> 00:17:35.282
Do you have a particular process?
00:17:35.282 --> 00:17:41.067
Do you like to storyboard or index things out, or do you just kind of write from the beginning to the end?
00:17:41.067 --> 00:17:42.200
How do you like to write a report?
00:17:43.296 --> 00:17:47.886
I usually do the report writing in a way that I interact with counsel.
00:17:47.886 --> 00:17:56.740
So sometimes counsel will send me something and I will be like, look, I don't like what it says here, I'm going to rewrite it.
00:17:56.740 --> 00:17:58.536
It doesn't fit.
00:17:58.536 --> 00:18:04.397
Or some of the you know attorneys make it more legalistic, I'm going to rewrite it.
00:18:04.397 --> 00:18:04.778
It doesn't fit.
00:18:04.778 --> 00:18:10.661
Or some of the attorneys make it more legalistic, so it needs to be more technical than legalistic and that's why I think it's a collaboration effort.
00:18:11.201 --> 00:18:23.526
The best reports are the ones that when I draft the first version, I send it back to counsel and we go back and forth maybe five, seven times and then we have a final version of that.
00:18:23.526 --> 00:18:26.126
We usually has to be signed.
00:18:26.126 --> 00:18:28.567
And I'm not kidding to all the experts out there.
00:18:28.567 --> 00:18:49.914
Don't be surprised if you're at 1153 pm, the day of the deadline, and you're still like signing and sending back the same report back to counsel, because it happens multiple times that it's like oh, maybe the client will be the actual patent owner or the client in particular cases.
00:18:49.914 --> 00:18:51.640
I don't like what is phrased here.
00:18:51.640 --> 00:18:52.623
Can you rephrase it?
00:18:52.623 --> 00:18:59.626
I don't like what it says here, or did you include this and this piece of evidence in your report?
00:18:59.626 --> 00:19:10.028
So you have all that back and forth with the client, which is the ultimate plaintiff or defendant, and the attorneys as well.
00:19:12.278 --> 00:19:13.603
Wow, that's really interesting.
00:19:13.603 --> 00:19:29.278
Before we wrap up, do you have, I don't know, a story or two that you can talk about that has kind of either influenced or informed the way in general that you are as an expert witness, either tactics that you use or things to avoid?
00:19:30.461 --> 00:19:34.132
Yes, well, I mean, I've been working on a project personally.
00:19:34.132 --> 00:19:36.156
It's called the P-Tab Expert.
00:19:36.156 --> 00:19:51.166
The P-Tab Expert is basically basically an ai tool, like a gpt tool that reads all the the pleadings, all the filings in the ipr uh, in the p-tab uh court.
00:19:51.166 --> 00:19:55.476
The p-tab is the, the patent trial and appeals board.
00:19:55.476 --> 00:20:00.138
For those who don't not familiar with, those records are open to the public.
00:20:00.138 --> 00:20:12.209
Anyone can download them and I have indexed a lot of the files from those records and the idea with the PTAB Expert is that you use machine learning and AI.
00:20:12.209 --> 00:20:16.711
Who will win or who may lose a particular challenge?
00:20:22.575 --> 00:20:26.103
Now, with this knowledge base that is learned, let's say, like you can, you can tell this this p-type expert, gpt.
00:20:26.103 --> 00:20:33.586
Let's say, hey, um, I have a case on this and that uh electronic device that is wearable, that is like I don't know.
00:20:33.586 --> 00:20:44.069
Um, it's a band that I wear, just like like the magic band from disney world, for example, and then it will try to identify okay, based on that, you need a wearable expert.
00:20:44.069 --> 00:20:45.421
Maybe this is the expert for you.
00:20:45.421 --> 00:21:12.486
This is the best firm for that particular case and this is the potential outcome based on data that is live from the PTAB database and ideally, it will be extended to use the federal circuit, but the PTAITA database is open to the public, so the AI and machine learning expertise is being gathered by myself right there and I will start using next year for some of the cases.
00:21:14.075 --> 00:21:16.423
Wow, that is absolutely fascinating stuff.
00:21:16.423 --> 00:21:19.496
Dr Hernandez, thank you so much for joining me here today.
00:21:20.618 --> 00:21:21.180
Thank you, Noah.
00:21:21.180 --> 00:21:22.303
Thank you so much for joining me here today.
00:21:22.303 --> 00:21:22.644
Thank you, Noah.
00:21:22.644 --> 00:21:23.204
Thank you for everything.
00:21:23.204 --> 00:21:23.826
Have a great afternoon.
00:21:24.268 --> 00:21:30.544
Happy New Year to everyone and thank you to our audience for joining me for another discussion at the Roundtable Cheers.
00:21:30.564 --> 00:21:32.431
Thank you for listening to our podcast Discussions at Roundtable.
00:21:32.431 --> 00:21:36.765
Our show notes are available on our website, roundtablegroupcom.
00:21:36.765 --> 00:21:41.605
Subscribe today on Apple Podcasts or your favorite listening apps.