Latest / Code Story | Startup Podcast for CTOs, CEOs and Technical Founders / S7 Bonus: The Death of Cookies: Engineering 99.5% Accurate Device Identification with Valentin Vasilyev of Fingerprint

Transcript

S7 Bonus: The Death of Cookies: Engineering 99.5% Accurate Device Identification with Valentin Vasilyev of Fingerprint · Code Story | Startup Podcast for CTOs, CEOs and Technical Founders

• 0:01
  [SPEAKER_00]: Getting validation from customers that you're building what they want is the number one priority for an early founder and I made that mistake of building customers didn't need.
• 0:12
  [SPEAKER_00]: But luckily, I didn't feel like it's the end of the world.
• 0:16
  [SPEAKER_00]: One of the values that we have is the fail learned grow.
• 0:20
  [SPEAKER_00]: So fail comes first.
• 0:22
  [SPEAKER_00]: So it's fine and our culture to fail then because you learn from that and then you grow the result.
• 0:27
  [SPEAKER_00]: So even though it was a failure to build certain things, or it was a mistake to build certain things so that we wouldn't sell, we'll learn from that in cruiser results, so it'll rise with our core value, which is a failure to learn grow.
• 0:38
  [SPEAKER_00]: My name is Valentino Vastiliev, I am the co-founder and CEO of ThinkerPrint.
• 0:48
  [SPEAKER_01]: This is Code Story, a podcast bringing you interviews with tech visionaries, who share in the critical moments of what it takes to change an industry and build and lead a team that has your back.
• 1:02
  [SPEAKER_01]: On your host, Snow Alapart, and today how Balantine Vasily took his open source pre-boat to the next level and created a VICEID platform to prevent fraud.
• 1:13
  [SPEAKER_00]: All this and more, on Code Story.
• 1:21
  [SPEAKER_01]: Valentine Vasiliev has always been interested in the open-source world.
• 1:25
  [SPEAKER_01]: His GitHub account was created way back in 2008 when they started.
• 1:28
  [SPEAKER_01]: He's always dreamt about contributing to Ruby on Rails since he developed with it in the past.
• 1:33
  [SPEAKER_01]: Outside of tech, he loves cycling, road, and mountain, and enjoys good coffee and online shooting games.
• 1:38
  [SPEAKER_01]: He's married with two kids and loves living in the Midwest, specifically Chicago.
• 1:43
  [SPEAKER_01]: Post-kidding his green card, Valentine wanted to experiment with the open-source library he created.
• 1:48
  [SPEAKER_01]: which its worth noting was starred 6,000 times.
• 1:52
  [SPEAKER_01]: He decided to keep his library but spin off a pro version which included a back-end.
• 1:56
  [SPEAKER_01]: Then he followed up with some landing pages and payment mechanisms.
• 2:00
  [SPEAKER_01]: And then began his focus on accuracy.
• 2:04
  [SPEAKER_01]: This is the creation story of FingerPrint.com.
• 2:10
  [SPEAKER_00]: The fingerprint.com is a new name and we were called fingerprintjs.com, the name sake of the library itself.
• 2:19
  [SPEAKER_00]: So we started as a library fingerprintjs.
• 2:21
  [SPEAKER_00]: The company was built on top of the open source and I'll give you a brief history or the timeline how the library, tiny library and GitHub turned into a company because I think it's an important step.
• 2:35
  [SPEAKER_00]: I left my job and I decided I want to experiment with the library and try to turn it into a SaaS business.
• 2:44
  [SPEAKER_00]: By then the library will start like 6,000 times I think, so it had this interaction and a lot of developers knew about it.
• 2:51
  [SPEAKER_00]: So I left the company and started just thinking about how can I turn it into a SaaS business.
• 2:58
  [SPEAKER_00]: I thought about multiple things like should it be?
• 3:01
  [SPEAKER_00]: with what the price model should be choose to customers by API calls or subscription.
• 3:07
  [SPEAKER_00]: So there are many questions and how to turn a JavaScript library into his house business.
• 3:11
  [SPEAKER_00]: So there were a lot of questions and I didn't know any answers.
• 3:14
  [SPEAKER_00]: So I just thought about it like kinds of different approaches.
• 3:19
  [SPEAKER_00]: So, in the end, I decided I will keep the library but make it a better version.
• 3:23
  [SPEAKER_00]: So, instead of calling it finger print jess, I called it finger print jess pro to signal that it's a professional level library, JavaScript that runs in the browser, and then I realized like it's so easy to see what's inside your JavaScript and when you minimize or you're not forsaken your JavaScript used to, it can very easily see what's going on there.
• 3:42
  [SPEAKER_00]: So, I needed a backend and on the backend where the cool stuff will happen.
• 3:47
  [SPEAKER_00]: and the client-site library will be just an agent that runs in the browser collects signals and then sends those signals to the back end.
• 3:54
  [SPEAKER_00]: And the back end is where it all the magic happens.
• 3:56
  [SPEAKER_00]: So I realized that and I built a better version of the client-site library called FingerPringes Pro and the back end to process all the data that the library sends and return back information that's valuable to customers.
• 4:08
  [SPEAKER_00]: That was one piece of the equation.
• 4:10
  [SPEAKER_00]: The second piece was obviously I needed a dashboard and the website.
• 4:14
  [SPEAKER_00]: And website is where the landing pages were and how you learned about it, what it could do, and the dashboard is how you sign up and I use your created card.
• 4:23
  [SPEAKER_00]: I think it's trip obviously because it was a natural choice.
• 4:28
  [SPEAKER_00]: And the idea about fingerprint, I wanted to create a product where each browser and each device on the planet can have a unique identifier.
• 4:37
  [SPEAKER_00]: and companies can use that identifier to understand when something suspicious or something nefarious is happening on their website or in their replication that was the main idea and the main challenge was how to generate that unique number because with open source library it tried to do that but it wasn't really successful the accuracy rate was about 60% and on iOS devices it was even lower around like 45%
• 5:03
  [SPEAKER_00]: So I realized I would need to make a lot of changes to the product, to make it more accurate, in the first place.
• 5:08
  [SPEAKER_00]: I focused on the accuracy and on marketing that accuracy to customers like the library has two times more signals that backends its very smart and sophisticated there is machine learning, etc.
• 5:18
  [SPEAKER_00]: And I used to work with machine learning before in the previous company.
• 5:21
  [SPEAKER_00]: So I
• 5:22
  [SPEAKER_00]: kind of combined everything and you about JavaScript type scripts and back in programming and machine learning in one product, thus the fingerprintages.com was born, a service that allows you to identify every browser on the planet you need with.
• 5:33
  [SPEAKER_00]: That's how it kind of turned from an open source library into a product.
• 5:38
  [SPEAKER_00]: We started with Michael Fondredan and he focused on selling and business development, I focused on building, so we had a very clear separation who does what, and I think this specialization helps us to move forward faster, especially in the early days,
• 5:53
  [SPEAKER_01]: Tell me about the MVP.
• 5:55
  [SPEAKER_01]: What you would consider the MVP.
• 5:56
  [SPEAKER_01]: Maybe that's the open source library, or maybe that's when you started to, you know, attempt to make the pro version and monetize it, I'll let you decide.
• 6:04
  [SPEAKER_01]: But tell me about that MVP.
• 6:06
  [SPEAKER_01]: You know, how long did it take you to bring it to life?
• 6:08
  [SPEAKER_01]: And what sort of tools did you use?
• 6:11
  [SPEAKER_00]: I started with an open source library on my hands and knowledge about general knowledge about this ecosystem or in general fraud detection and browser fingerprinting as a subject.
• 6:22
  [SPEAKER_00]: When building the first amp, like version the MVP, I used Ruby on Rails for the backend because that's what I knew as a Ruby developer, but then I realized like early on, I know this thing is gonna
• 6:33
  [SPEAKER_00]: require high scale because this is an infrastructure provider.
• 6:38
  [SPEAKER_00]: So it will run on large web applications and websites.
• 6:42
  [SPEAKER_00]: It needs to be really scalable and robust.
• 6:45
  [SPEAKER_00]: So very early like maybe after two or three weeks of working on the MVP, I switch to goal link for the backend.
• 6:51
  [SPEAKER_00]: And at that time I didn't know any goal-ling.
• 6:53
  [SPEAKER_00]: I knew that the language existed.
• 6:55
  [SPEAKER_00]: I, you Google created it to make it simple language easy to get started for their engineers.
• 7:01
  [SPEAKER_00]: And they kind of used this for all their products and all their services.
• 7:04
  [SPEAKER_00]: I was surprised how productive I immediately became with the goal-ling.
• 7:08
  [SPEAKER_00]: client site was built in TypeScript.
• 7:11
  [SPEAKER_00]: I'm a fan of TypeScript because in the past I used to be a C-sharp developer and TypeScript in C-sharp or really close other types of type system all of that.
• 7:20
  [SPEAKER_00]: So the MVP was built TypeScript on the front, goal link on the back, and that's what I actually deployed to AWS the last BeanSock.
• 7:29
  [SPEAKER_00]: On the cloud, I use DynamoDB because I wanted to try NoSQL key value database, and I will use a web application and a worker.
• 7:38
  [SPEAKER_00]: And I use SQL to connect it to.
• 7:40
  [SPEAKER_00]: Essentially, when you got an API request, it is processed by a web server.
• 7:44
  [SPEAKER_00]: It puts it on the SQL queue, and then the worker puts it up, and I put scale.
• 7:49
  [SPEAKER_00]: workers independently from the front end and it was very scalable from the start.
• 7:53
  [SPEAKER_00]: So I kind of violated that mantra for a start-up founders don't scale per material, I did scale per material, I made that mistake as like other people did.
• 8:05
  [SPEAKER_00]: It took me about two months to build the first version, I think.
• 8:09
  [SPEAKER_00]: I started in April of 2019, and around May and of May already had the first version that ran in production, and I got the first traffic from a company who was interested to try it out.
• 8:21
  [SPEAKER_00]: So two months from start until, until first traffic from a real customer, that was the timeline.
• 8:28
  [SPEAKER_00]: It wasn't stable or anything like that, but it did work, and customers kept using that, which was surprising, and they kept asking questions about it, and I kept like fixing things, and pushed a lot of commits every day deployed to production multiple times per day.
• 8:44
  [SPEAKER_00]: So it was a very intense period.
• 8:47
  [SPEAKER_01]: Okay, so from that point you've got the MVP.
• 8:50
  [SPEAKER_01]: How did you progress it forward, you know, into, you know, fingerprint.com today and how are you maturing the product?
• 8:57
  [SPEAKER_01]: And I think to wrap it up box a little bit, make a little more sense here.
• 9:00
  [SPEAKER_01]: How are you building your roadmap?
• 9:02
  [SPEAKER_01]: How are you deciding, okay, this is the next most important thing to build or to address with fingerprint.com.
• 9:09
  [SPEAKER_00]: I have this idea in my head, what it needs to be, what it needs to become as a final product.
• 9:15
  [SPEAKER_00]: And I want every business on the planet, run in a safe environment.
• 9:21
  [SPEAKER_00]: And to do that, they need to know when a bad user is trying to use their system.
• 9:27
  [SPEAKER_00]: And this general idea impacts significantly how I think about it, like every feature that I think or every feature that we hear from customers,
• 9:39
  [SPEAKER_00]: When we discuss the roadmap, who have a product, you know, the roadmap is something that they work on and prioritize.
• 9:47
  [SPEAKER_00]: But it's always a combination of talking to customers, data from customers, and what customers want.
• 9:53
  [SPEAKER_00]: But then on top of that, there is a big filter of this original idea of what it should be.
• 9:59
  [SPEAKER_00]: So, and the original ideas, every business should have a system where they can build
• 10:07
  [SPEAKER_00]: and they know more about bad users, they know more about good users so that they could protect their data, their accounts, and the companies using finger print could also grow their business and grow their accounts and grow usage.
• 10:20
  [SPEAKER_00]: So it's a two-fold things, it's knowing more about bad users and protecting your data from them and knowing more about good users to reduce friction and help you grow your revenue.
• 10:30
  [SPEAKER_00]: So it's a two-fold approach that kind of dictates
• 10:35
  [SPEAKER_00]: very careful about not working on something that does not align with that original notion that existed from the very beginning.
• 10:43
  [SPEAKER_00]: Going from an MVP to this stage was a lot of work because MVP was built in two months, but then it was zero revenue.
• 10:50
  [SPEAKER_00]: Then there was first revenue, first companies buying the product.
• 10:54
  [SPEAKER_00]: And in several more months, I got to about 100,000 ARR, and then we started scaling more aggressively with my co-founder.
• 11:01
  [SPEAKER_00]: We invested a lot more effort into
• 11:04
  [SPEAKER_00]: selling in the correct because I didn't know how to sell and we started to build processes etc and attempted to grow a revenue a lot more and try to sell larger contracts to larger businesses.
• 11:17
  [SPEAKER_00]: In a couple more months, I grew the revenue to like I think two or three hundred thousand and that's when we started thinking about raising venture capital.
• 11:26
  [SPEAKER_00]: So we closed the C-drowned in August of 2020 and that's when we became a startup.
• 11:32
  [SPEAKER_00]: So before that it was like just to guys try to sell and try to build.
• 11:37
  [SPEAKER_00]: But in July of 2020 we closed the first round and we started hiring and then it's more or less like standard story, multiple rounds of funding.
• 11:46
  [SPEAKER_00]: Now the company is 110 people.
• 11:49
  [SPEAKER_01]: Okay, let's switch to team then.
• 11:51
  [SPEAKER_01]: So you mentioned your, you know, a much bigger team obviously than two guys in the beginning.
• 11:55
  [SPEAKER_01]: How did you go about building that team?
• 11:56
  [SPEAKER_01]: And what did you look for in those people to indicate that they were the winning horses to join you?
• 12:02
  [SPEAKER_00]: Building a team as I think the hardest part is it takes so much effort, so much time.
• 12:07
  [SPEAKER_00]: Just a few, an example.
• 12:09
  [SPEAKER_00]: Our first engineer, I interviewed 70 other engineers before we gave an offer to the first person.
• 12:14
  [SPEAKER_00]: It's just to give you the scale of how much work it is to find early people, so 70 interviews to give an offer to the first chosen one that will be your first engineer.
• 12:26
  [SPEAKER_00]: And that kind of continued like that, a lot of interviews just to move forward with offers like really, really slowly, be very careful in offering like people to work with you.
• 12:39
  [SPEAKER_00]: Logistically, we focus on hiring outside of the US.
• 12:43
  [SPEAKER_00]: Our first engineer is from Russia and he referred a lot of great engineers and This is how the core of the early the early core was was built then we expanded into Ukraine Then we expanded into Czechia Slovakia and we kind of continued hiring in that aerial Central and Eastern Europe and then another cluster of engineers appeared and
• 13:05
  [SPEAKER_00]: I knew a person from Argentina and he recommended a great developer.
• 13:09
  [SPEAKER_00]: We hired that engineer as well, very early.
• 13:12
  [SPEAKER_00]: I think he was engineer number two.
• 13:13
  [SPEAKER_00]: So in 2020 we hired two engineers to first engineer as one was from Russia and the other one was from Argentina and both helped us to grow like the engineering teams around them in those respective areas.
• 13:27
  [SPEAKER_00]: and we expanded like I mentioned to Ukraine and Slovakia, etc.
• 13:30
  [SPEAKER_00]: about in Argentina we expanded into Uruguay and Brazil etc.
• 13:35
  [SPEAKER_00]: so we ended up the year 20 to point it with two groups of engineers one in Central Eastern Europe and another one in South America.
• 13:43
  [SPEAKER_00]: In 2021, we continued hiring and expanded into other areas North America, Canada, United States, Asia, and the biggest hiring was in 2022.
• 13:54
  [SPEAKER_00]: The company, I think, started 2022-27 people, and we ended the year as like 100 people.
• 14:02
  [SPEAKER_00]: So, we almost forexed the headcount in just
• 14:12
  [SPEAKER_00]: interviewing carefully, interviewing a lot, a lot of interviews, being on titles, being with your offers, and pay a lot of attention to the culture, like a lot of things are important in addition to engineering skills, and I think it's paid off now that the team is 45 engineers, and to we, we have built this culture of like being helpful to each other, having a low ego environment.
• 14:37
  [SPEAKER_00]: Of course, good engineering skills, but all the soft skills that we were looking for, they helped tremendously to shape that culture that we are now working in.
• 14:50
  [SPEAKER_01]: Okay, let's switch to scalability then.
• 14:53
  [SPEAKER_01]: So was this built to scale efficiently from day one or have you been fighting this as you grow in gain traction?
• 15:00
  [SPEAKER_01]: And this could be, you know,
• 15:05
  [SPEAKER_00]: First of all, it was built to scale from day one, and now I think it might have been a mistake to focus so much on scaling in the early days.
• 15:14
  [SPEAKER_00]: When I was building a VPI focused a lot of like how many requests per second this thing can handle or how can I scale it easily?
• 15:21
  [SPEAKER_00]: And I spent a lot of time thinking about it, but in the hindsight I shouldn't have, and I should have focused in business development more of course.
• 15:28
  [SPEAKER_00]: And I think this is a very fairly typical mistake.
• 15:32
  [SPEAKER_00]: Scaling was easy because of that.
• 15:35
  [SPEAKER_00]: Once we did get the traction from customers, it was fairly easy for us to scale because the system was already built to be scalable.
• 15:42
  [SPEAKER_00]: It was a worker Q-based system and to scale it all we had to do is just to add or note or EC2 instances and with easily double or triple or quadruple the number of instances and it could just start handling
• 16:01
  [SPEAKER_00]: On one hand, it was a mistake that I focused on scalability so early and maybe didn't grow as fast, revenue and the business as I could have, but on the other hand, once we did get the traction, did get the first big customers.
• 16:14
  [SPEAKER_00]: It was very easy for us to scale that system by just adding, throwing in more easy to instances, adjusting our auto-scaling rules, adjusting our dynamo did be reserved capacity.
• 16:24
  [SPEAKER_00]: That's all we did to scale.
• 16:26
  [SPEAKER_00]: we kind of continue doing that to this day.
• 16:29
  [SPEAKER_00]: The system is more or less built in the same way, just scale from 500 requests per second to 1000 requests per second, we just need to wait and the other scaling rules will double instances or just triemptively double the number of instances ourselves.
• 16:43
  [SPEAKER_00]: It's very easy now.
• 16:46
  [SPEAKER_01]: So as you step out on the balcony and you look across all that you've built,
• 16:53
  [SPEAKER_00]: I am proud of the team, I am proud of the culture, like this idea that people enjoy working together, they enjoy working on the product together, and they love helping each other.
• 17:04
  [SPEAKER_00]: It's not an environment where you feel like not empowered because no one wants to help you.
• 17:09
  [SPEAKER_00]: When you come, you join, as a as a new engineer, I got that feedback a lot of times, like I was surprised how exciting and willing to help other engineers to work when I joined.
• 17:18
  [SPEAKER_00]: how welcoming they were.
• 17:19
  [SPEAKER_00]: And to do this level of culture I think is the biggest achievement because that's what allows you to keep growing as a business without suffering, stream difficulties of scaling because people just are not excited to work together.
• 17:33
  [SPEAKER_00]: So culture is number one definitely.
• 17:35
  [SPEAKER_00]: Technology is number two.
• 17:37
  [SPEAKER_00]: Ingerpring provides an amazing technology that allows you to identify devices and browsers.
• 17:42
  [SPEAKER_00]: We went from a JavaScript library initially to professional-level JavaScript library then a SaaS business that's getting identified browsers.
• 17:49
  [SPEAKER_00]: Then after we realized we can identify browsers really well, we realized we needed to add mobile device supports that we added.
• 17:57
  [SPEAKER_00]: and draw it than iOS.
• 17:59
  [SPEAKER_00]: Then we started making it easier in we built a developer experience team with a focus of making it easy for developers to start using the product so we built around 15 frameworks and as the case on libraries any major library framework we have an integration for the react and
• 18:26
  [SPEAKER_00]: We built all kinds of integrations and realized companies needed to plug into their stacks like so we built a cloud flaring integration that really helps to get started quickly with the best practices already packaged in the cloud flaring integration.
• 18:39
  [SPEAKER_00]: And then later we added cloud front integration with the same idea.
• 18:42
  [SPEAKER_00]: Let's give our customers a package once you enable that package.
• 18:47
  [SPEAKER_00]: You have all the best practices in one place.
• 18:49
  [SPEAKER_00]: It's about the privacy security at blocker protection.
• 18:53
  [SPEAKER_00]: Identification accuracy performance, caching, all of that in one package, specifically for your favorite web framework.
• 19:01
  [SPEAKER_00]: For example, you could say, I'm using Python Flask on the back end, and I'm using React on the front end, and I'm running this in Cloud Flare.
• 19:08
  [SPEAKER_00]: We give you the whole package.
• 19:09
  [SPEAKER_00]: We give you all three, and all three are filled with best practices that you can get started really easily.
• 19:15
  [SPEAKER_00]: So, the developer experience team, they built all that.
• 19:18
  [SPEAKER_00]: that allows our customers to start quickly and get the best results right away without having to tune their integration or tweak some settings in slow-ready prepackaged.
• 19:27
  [SPEAKER_00]: So I think that technology is number two for me in terms of what I'm proud of because not only is the API self is pretty powerful and covers every platform and every integration but it allows you
• 19:39
  [SPEAKER_00]: For businesses and developers to get started so quickly and have all the best practices already there I think it's powerful and that's why like the product sells well and customers loved that would be in my number two Okay, let's flip the script a little bit.
• 19:55
  [SPEAKER_01]: Tell me about a mistake you made and how you and your team responded to it
• 20:00
  [SPEAKER_00]: In general, I would say mistakes that I made is building something no one used after that, like thinking we need that feature or that product or something like that.
• 20:10
  [SPEAKER_00]: And then just we didn't need that we had to table that idea or table that code or this repository.
• 20:17
  [SPEAKER_00]: It happened, it didn't happen a lot, it happened like one or two times in the early days.
• 20:20
  [SPEAKER_00]: But I think getting validation from customers that you're building what they want is the number one priority for an early founder.
• 20:29
  [SPEAKER_00]: And I made that mistake of building customers didn't need.
• 20:33
  [SPEAKER_00]: But luckily, I didn't feel like it's the end of the world.
• 20:42
  [SPEAKER_00]: One of the values that we have is the fail learned grow, so fail comes first, so it's fine and our culture to fail, then because you learn from that and then you grow the result.
• 20:52
  [SPEAKER_00]: So even though it was a failure to build certain things or it was a mistake to build certain things that we didn't need or couldn't sell, we'll learn from that in cruiser results, so it applies with our core value, which is fail learn grow.
• 21:06
  [SPEAKER_01]: So what does the future look like for fingerprint.com the product and for your team?
• 21:12
  [SPEAKER_00]: We started with building this platform of identification device identity platform.
• 21:18
  [SPEAKER_00]: Now how do we use that?
• 21:19
  [SPEAKER_00]: How do we make it valuable for customers around the world?
• 21:24
  [SPEAKER_00]: Today we are a signal provider.
• 21:25
  [SPEAKER_00]: So we give businesses a signal and this main signal that we give is that device had year browser ID.
• 21:31
  [SPEAKER_00]: But we don't really tell you when you should block a user because this user is trying to take over someone's account or when you should pass the user through, when it's safe, when it's unsafe, what's the risk, etc.
• 21:43
  [SPEAKER_00]: So the next step for us is to take that identity capability and the identity platform and turn it into a decision making engine customers can utilize without having to write any code.
• 21:56
  [SPEAKER_00]: So what we the next challenge for us is to take the
• 22:00
  [SPEAKER_00]: Then build a rule-engineer decision making engine around that to help businesses build their own rules that are domain-specific.
• 22:08
  [SPEAKER_00]: And once they build those rules, those domain-specific decision-making kind of algorithms on our platform will know which these varieties which device IDs are high quality, good users, and will know which ones are fraudulent or nefarious bad users.
• 22:25
  [SPEAKER_00]: Currently, our keeping track of around several hundred million device IDs per month.
• 22:31
  [SPEAKER_00]: So we have a pretty good coverage in terms of how many devices we cover with our product.
• 22:38
  [SPEAKER_00]: But what we lack is that decision making ability and we are building that today to enable businesses to use this engine and get results from us, for example.
• 22:48
  [SPEAKER_00]: A visitor ID from Android device
• 22:53
  [SPEAKER_00]: geolocation, for example, and we can detect that, and we can say this visituriety is a geolocation school visituriety, so treated with caution.
• 23:02
  [SPEAKER_00]: This browser visituriety tried to log into 500 accounts in the last three hours, so this is clearly an account take over a
• 23:15
  [SPEAKER_00]: So the next thing for us is to take those rules that we built on the rule engine and mark or annotate every single visitor ID on the planet that we collect with valuable business risk and reputation information.
• 23:28
  [SPEAKER_00]: And then we share that information with our customers.
• 23:31
  [SPEAKER_00]: So one big customer maybe submits all the bad visitor IDs to us because they trigger some rule about a counttick
• 23:40
  [SPEAKER_00]: Then we can take those annotations about that bad visitor idea and share the fact that this is a bad visitor idea with other businesses.
• 23:48
  [SPEAKER_00]: And this is how it becomes extremely useful for everyone that the reputation or the fraud or the resulated flows and propagates globally from everyone to everyone.
• 23:57
  [SPEAKER_00]: And there is no personal ID and file information in that information.
• 24:01
  [SPEAKER_00]: It's just a fact that there was some and if there is activity or maybe a contact or attempt activity, and we could share that information with everyone, making it extremely useful for everyone.
• 24:11
  [SPEAKER_00]: So to summarize, we built an identity platform.
• 24:15
  [SPEAKER_00]: Now we're building a decision engine on top of the platform that will enable every business on the planet to utilize and use all the good and bad information the annotations and the risk about each visitor ID that we collect.
• 24:28
  [SPEAKER_01]: Let's switch to you who influences the way that you work.
• 24:32
  [SPEAKER_01]: Name someone or many persons or something you look up to and why.
• 24:37
  [SPEAKER_00]: I thought about it like several days ago and I thought that I had to don't try to influence by people.
• 24:42
  [SPEAKER_00]: I tried to be influenced only by ideas.
• 24:44
  [SPEAKER_00]: So if I see someone I like, even in that case I say, no, I'm not going to do what this person does.
• 24:50
  [SPEAKER_00]: What I like, why I like this person is because of this idea or this.
• 24:55
  [SPEAKER_00]: concept that I admire maybe or think is very cool.
• 25:00
  [SPEAKER_00]: So I take that concept and apply it to what I do.
• 25:03
  [SPEAKER_00]: So I cannot name names because I don't try to follow people but as for concepts like I do try to distill things that make certain ideas cool and try to adopt them and brace them and use them.
• 25:19
  [SPEAKER_01]: Valentin last question, so you're getting on a plane and you're sitting next to a young entrepreneur who's built the next big thing.
• 25:25
  [SPEAKER_01]: They're jazzed about how they get me to show it off to the world and can't wait to show it off to you right there on the plane.
• 25:30
  [SPEAKER_01]: What advice do you give that person having got down this road a bit?
• 25:34
  [SPEAKER_00]: In general, I think advice is, don't try to give advice unless someone asks for it, if someone did ask for my advice, I would just tell a founder to trust their instincts, to push forward regardless of what everyone says about, like,
• 25:49
  [SPEAKER_00]: Whether it's going to be successful, whether this idea is good or bad, just don't listen to people, push forward, trust your instincts, and read less about like stuff or startups on Reddit or indie hackers.
• 26:01
  [SPEAKER_00]: Those platforms are valuable, but it's important not to overdo that I think, and trust you got more, then trust the advice of other people.
• 26:10
  [SPEAKER_00]: So, that actually means don't listen to my advice too, because I'm giving an advice and you don't need to trust that as well.
• 26:16
  [SPEAKER_00]: Just do what you think is right, because Founder Instincts sometimes most often is the best way to move forward to decide about something.
• 26:23
  [SPEAKER_00]: Very often you don't have data, if you're doing AB tests, you think you will have data but you don't, it's the same, it's not inclusive, and be at the same.
• 26:32
  [SPEAKER_00]: Just go with your Instinct, like if you like, beaches go with P, if you're not sure, just sleep on it, next morning decide what you want to do because you like.
• 26:42
  [SPEAKER_00]: I watched this movie like put the third time I guess recently it's called Prometheus about the alien universe and one scientist like was pursuing crazy ideas about the origin of life and one person asked the scientists how do you know this and the scientists answered I don't know but this is what I choose to believe
• 27:02
  [SPEAKER_00]: And I think this is a very powerful framework of thinking.
• 27:04
  [SPEAKER_00]: You do not know a lot of things, but you need to decide what you choose to believe.
• 27:09
  [SPEAKER_00]: And when you do decide what you choose to believe, a lot of things get easier, become easier for you.
• 27:14
  [SPEAKER_00]: Don't listen to other people, just do what you, you believe in, and I think it will work out well.
• 27:19
  [SPEAKER_01]: The Fantastic Advice.
• 27:20
  [SPEAKER_01]: Valentina really appreciate you being on the show today.
• 27:22
  [SPEAKER_01]: Thank you for telling the creation story of fingerprint.com.
• 27:25
  [SPEAKER_00]: Thanks for having me.
• 27:28
  [SPEAKER_00]: And this concludes another chapter of Code Story.
• 27:34
  [SPEAKER_01]: Code Story is hosted and produced by Noah Labhart.
• 27:37
  [SPEAKER_01]: Be sure to subscribe on Apple Podcast, Spotify, or the podcasting app at your choice.
• 27:42
  [SPEAKER_01]: And when you get a chance, leave us a review.
• 27:45
  [SPEAKER_01]: Both things help us out tremendously.
• 27:49
  [SPEAKER_01]: And thanks again for listening.