The following interview is a conversation we had with Ravin Thambapillai, CEO and Co-Founder of Credal, on our podcast Category Visionaries. You can view the full episode here: $5 Million Raised to Build the Future of AI Security.

 

Ravin Thambapillai
Thank you so much for having me excited to talk. 

Brett
Not a problem. I’m super excited as well. And I’d love to just kick off with a quick summary of who you are and a bit more about your background. 

Ravin Thambapillai
Yeah, so, as with any person, right, it’s really hard to summarize all of me in like 1 minute, but I will do my best to focus on the pits actually relevant to the audience. I think probably the most interesting or most relevant part of my background is just, yeah, I’m a person who’s always looking for the most impactful thing I can be doing. I care like a lot about just being very laser focused on the thing that I think is most impactful at any given point in time. And honestly, I’m so dogmatic about that it’s gotten me into a lot of trouble over the years. I grew up in a very traditional, risk averse immigrant household. My parents actually fled the war in Sri Lanka. 

Ravin Thambapillai
And so given that they lived through this like insane upheaval in their life, they were like very much not risk takers. Right? So we ended up in England and I managed to get myself a scholarship place sort of insanely stodgy british boarding school. And it was kind of comically rigid. It had all these sort of absurd rules and etiquette stuff. And honestly, none of that stuff ever made any sense to me at all. I was there for like twelve years and I never figured any of it out. So I pretty much grew up in this environment where literally the best thing you could possibly do was not think at all about what you were doing and just kind of blindly follow whatever arbitrary the rule of the day there was. Right. And obviously that was kind of the opposite of me. 

Ravin Thambapillai
I was continuously in all sorts of trouble for resisting these arbitrary rules. And in fact, I think I was the only scholarship kid in the school who got suspended multiple times. And then they kicked out of the school. And in fact, actually, when it came time to apply to university, my professors at Oxford ended up telling me that I was the only kid they’d ever seen whose recommendation letter was explicitly saying, do not admit this student, you’re going to regret it. Ignore the fact they have great grades, they’re like a bad kid, and they have a huge problem with authority, and you should just not let this kid in. And I know somehow they let me in anyway. 

Ravin Thambapillai
And while it was a lot better than the school in terms of being able to explore and learn what I wanted to learn, it did have its own strange norms as well. Most kids at Oxford at the time did not see it as a way to learn about important, impactful problems. They saw it as a way to get a fast track to high prestige investment banking jobs. And I think back then in the UK, in the late two thousands, literally the highest prestige would be to go to, say, McKinsey, where you could sell opioids to rural America, or Goldman Sachs, where you could grift the malaysian people out of their national wealth, right? And these were literally the highest prestige jobs you could get. Drug dealer or international scale grifter. Right? And yeah, I’m obviously being a little bit hyperbolic, but. 

Ravin Thambapillai
But also, like, there’s a kind of truth in all that. So this wasn’t me, obviously. So I did wind up at Google hoping that building stuff would be a bit more impactful. In some ways it was better. But ultimately I wound up joining a fintech startup called go cardless, which was definitely, like, more impactful and something I was more excited about. And then eventually wound up a palantir for eight years where I really actually get access to these amazing problems. I worked on rehousing syrian refugees, I worked on keeping America’s hospitals afloat during the pandemic. But, you know, I think I eventually realized that ultimately, if I really wanted to be dogmatic about working on what I felt was the most impactful problem in the world, the best way to do that was at a startup. 

Ravin Thambapillai
And really that’s how I ended up starting Credal. 

Brett
Why do you think Oxford let you in? 

Ravin Thambapillai
Yeah, great question. Well, a lot of it is down to this incredible professor. I had Walter, who I spoke to him about it afterwards, and he got this recommendation letter, and he was also a bit of a rebellious character as well. And so he was kind of like, well, I have to meet this guy, right? Like, this is such a crazy reference. I have to meet this person. And so they brought me in and interviewed me, and, you know, my recollection of that interview was that I did a terrible job, but I suppose their recollection of it was a little bit more charitable than that. Yeah, we did have a really interesting conversation about the role of power international relations. It was pretty fun, and, you know, I made all sorts of silly mistakes, but it was still a good conversation. 

Ravin Thambapillai
And I think that was an interesting and weird enough profile for them that they were kind of excited to just see who I turned out to be. Was I as weird and crazy as. As the school was saying? 

Brett
What was it like then at palantirs? He spent eight years there. That’s a long time. That’s a good chunk of your career. I’m sure you learned a ton. What would be some of those big takeaways that you walked away with? 

Ravin Thambapillai
Yeah, great question. Well, I think one of the things that was super interesting about Palantir when I joined it, so I joined in the London office in 2015, and at the time, I’m pretty sure this is accurate. This is obviously going back almost ten years. So I might be getting the numbers a little bit wrong, but I think we had, like, three customers in Europe, in the commercial world, and you could count every single person in Europe who had actually used palantir. A human being could count all of those people. Right? And if you were in the London office, you knew the names of all of the users in all of Europe, full stop. And that was kind of pretty cool in some ways. 

Ravin Thambapillai
Like, it’s quite analogous to being at a startup where, especially the early days of a startup, where you really do understand, like, the ins and outs of your users lives. You know, are they married? Do they have kids? Do they like to their job? Do they hate their job? You really get to know them very deeply, and that was, like, a very powerful lesson. I think there was one other thing that was really important about, like, my time at Palantir, which was when I was at Palantir, right. And it seems so obvious that it would be successful in hindsight, but when I was there, it really was not obvious at all. And, in fact, I remember at the end of 2016, the company had grown, like, I think, 3% in the whole year, and that was kind of a disaster, right? 

Ravin Thambapillai
Everyone was like, oh, my God, the company screwed. A lot of people left. I remember Carl, like, the CEO, he’s like a bit of a meme generator now. At the end of year, talk he sort of put up this graph and he couldn’t really have a graph of revenue because the revenue was flat. Right. So instead, the y axis on this graph just said, charisma. And he was like, charisma has cut off. Right. That’s kind of ridiculous thing that he would do. And were all like, what the hell is this? 

Ravin Thambapillai
But coming through that period and sticking with Palanza, because I did really believe in its mission and seeing it get through that sort of very difficult period and become the insane behemoth that it is today, I think that just taught me, hey, successful companies go through hard patches, and the ones that succeed are the ones that find a way out and don’t give up. And I think that was a very important lesson for me, ultimately. 

Brett
So as you made your way through Oxford, then went to go cardless, then you were at Palantir. In the back of your head, did you have the idea that someday you were going to start your own tech company, or when did that idea really start to develop? 

Ravin Thambapillai
Yeah, it’s a good question. I mean, I think when you’re at a startup, everyone is talking about, like, startups, right? And so the notion of starting a startup was not that alien to me, which I will say, like, again, at the time in the UK, in the late two thousands, early 2010s, starting a startup was not really something that british people. Right. It was kind of weird. I think when I was a God, cardless, there were essentially two interesting startups in all of the UK. And so it was a bit weird. And, you know, my very traditional risk averse parents, like, this is an awful idea. Why on earth would you imagine contemplate doing something like that? And so the more people tell you that if you’re like a very contrarian, somewhat rebellious person, the more drawn to that idea you are. 

Ravin Thambapillai
But to be honest with you, I think the reason that I didn’t do it for so long was because I did really care about working on extremely impactful problems. And I just didn’t feel like I had the ability by myself when I was like, 21, 22, 23 to really go away and make a big dent on what I thought was one of the world’s most important problems. These startups go Kabla St. Palantir, and obviously Google was not a startup, but these companies, I felt like they gave me a platform to do that. It was only when I’d really felt like I’d learned enough and I, like, deeply understood the problem spaces that I cared about, that I was okay. Like, it’s time. Now let’s go and actually tackle this myself. 

Brett
Is there a specific moment that you can recall in looks like, what, fall of 22, or maybe late 22, that you decided to go and start your own company? 

Ravin Thambapillai
Yeah, actually, it was very concrete. So I was in New York, and Jack. Jack’s my Co-Founder, by the way. Jack and I, were both working at Palantir at that point in these insanely high security environments. So Jack was at the Department of Defense, and he was helping them deploy these, like, combat computer vision models. And at the same time, I was at the Department of Health, right, and I was helping them deploy models to prevent America’s hospitals from overflowing from COVID And were both playing with these generative models and seeing how insanely powerful they were, but also seeing at the same time, like, how insanely difficult it would be to get these large, security sensitive institutions like trust and deploy them. And so we had dinner at this vietnamese place in Chelsea called Omai, which, very sadly, is actually closed now. 

Ravin Thambapillai
But during that conversation, we just both kind of simultaneously realized, holy cow, this is going to be a massive problem for every industry that touches sensitive data, which is basically every important industry in the world. And were seeing what we thought was, like, two or three years before the rest of the world, but turned out to be much less than that. But were seeing how insanely powerful the models were. And were, like, getting this technology to be safe and usable on the problems that matter the most to, like, human welfare. That’s just, like, an insanely valuable problem. We have to go and tackle it, because just the upside is so massive to humanity if we do it well. Now, that was in August of 2022 that we had that conversation. We thought were two or three years ahead of the world waking up. 

Ravin Thambapillai
But what happened was, two or three months later, chat, GPT was released, and we turned out to be off by two or three years, ironically enough. But, yeah, that was the specific moment, and it was very concrete, a very specific conversation we had. 

Brett
Take us back to those early days, then. What did the first, let’s say, 90 days look like? 

Ravin Thambapillai
It’s a lot of chaos, right? I think Jack and I were very determined that we wouldn’t be these, like, naive engineers who would go off and, like, build a bunch of software that nobody actually wanted, and then we would, like, struggle to do anything with it. So our goal was very much, okay, in the first seven days, we’ll build a demo. And in that demo, once we have that demo, we’ll go and start shipping it around, as many people that we can get to look at our demo and try and sell it to them. Obviously, the goal will be like, once you make the sale, then you actually go and turn the demo into something that someone can actually use. And we did actually get a handful of customers. Obviously, these are not massive contracts, but we did actually sell this very initial product. 

Ravin Thambapillai
The initial product that we ended up with was actually super different to what we wound up with today, because although we cottoned onto how important security was going to be, for some kind of silly reason, we didn’t actually understand that the right thing to do was build a security product. What we thought we would have to do is just build some very specific AI application and then bake a bunch of security features into that application, and that would help us sell here. We were selling this AI application out to a couple of bunch of companies, and they would kind of say, like, oh, you know, this product is sort of interesting, but the security features are really cool. But I’m sort of interested enough in you guys that like, you know, I’ll give the product a whirl. 

Ravin Thambapillai
And there was this one moment, actually, that was really difficult, where we got this pilot with this company called Latchel, which is this very innovative property management firms, about 100, 200 people, or 100 to 200 people. And they were using the product which was at the time like an AI chief of staff. It would look at all of your data and figure out what was going on in your meeting and help brief executives for exec meetings or hands. And he took a video of this exec meeting at the end of the pilot, and he asked everyone in the room, tell me, how much would you pay for this piece of software? This is a 100 to 200 person company. So were thinking we would charge $2,000 a month, something like that. And that’s what were hoping that they would say. 

Ravin Thambapillai
And so we watched this recording, and he goes around the room, and the execs are like, well, maybe I’d pay like $20 a month. Maybe I’d pay $10 a month. And you watch that, especially as a Founder of a company, and you realize, oh, no, something has gone horribly wrong. We’re off by two orders of magnitude on the price that we need here. And that was a pretty intense moment that happened. I don’t know exactly when, but like two or three months into the company. And at the end of that, Jack and I sat down, and were like, what the hell do we do? And just very shortly after that meeting, actually, we had a meeting with our advisors at YC, and I was writing up the notes for that. 

Ravin Thambapillai
And as I was writing up the notes, it was just like super obvious to me that this business would not work, right? Like, there was no way that you could sell a product for like ten to $20 a month to a 200 person company and turn that into a multi billion dollar business. So it was kind of gut wrenching and pretty hard. But in the end, it proved to be like this massive gift, right? Because I think if they’d said, oh, wed spend $200 a month or $300 a month, maybe we would have continued barking down that wrong tree or trying to pursue this dead end. And actually, the fact that were so far off from what could plausibly work actually accelerated our realization that, okay, this was not going to work. We should do something completely different. 

Ravin Thambapillai
Then we sat down and were like, this process took a little bit longer, but we sat down and were like, what have we learned in this failure? I think the thing that we realized was, look, our original thesis that the security problem is really big. That’s come out time and time again in the sales process. What has not come out is that anyone cares about an AI chief of staff. Let’s just focus on the thing that actually we wanted to do in the first place, which is figure out how to help enterprises trust this technology. And I think it was when we realized that we swept out a lot of the craft in the product and just focused on that. And thats when our business really actually started succeeding and blowing up and people really understood what were selling. 

Ravin Thambapillai
And I think that was a really pivotal moment. But honestly, the 90 days to get there was were all over the place making every mistake, and it was hard. 

Brett
When did you get your first paying customer then, to that AI security product? 

Ravin Thambapillai
Yeah, so the AI security product, our first paying customer, was actually checker. And it started out because I emailed the Ciso of Checker and on his LinkedIn, like, he kind of mentioned that he spoke Klingon, right? Or he was like, interested in Klingon. And so this like outreach email I sent in Klingon kind of being like, well, you know, let’s see what happens. And he picked it up, but he was like, actually, it’s funny that you should mention this. This was, I think, February of 2023. It’s funny that you should mention this. Like, we’re actually thinking about building this exact thing in house ourselves right now, but I think we’d sort of prefer to buy it from someone. So if you have that product, great. 

Ravin Thambapillai
So then Jack and I panicked because we didn’t even have a demo for this product at that point and we like scrambled to create a demo. We met with the checker team, like literally the next day because they were actually pretty urgent about it. And we met with Michael, the CISO, as well as a few other people from around the business. And it was just really clear that they wanted to get this done. We ended up signing like a three month pilot agreement with the idea being that over the course of those three months, we’d like actually build the product, right. And then at the end of that three month period, it converted to like a really awesome six figure enterprise deal. And that was kind of, I think, when we realized, ok, we had something valuable here. 

Ravin Thambapillai
We had something people wanted and something that people really wanted to pay for. It just kind of kept going from there. 

Brett
How many other emails did you send out like that? Obviously not in Klingon, but just emails. An extreme amount of personalization and time and care. How many did you send out? 

Ravin Thambapillai
I mean, if we’re just talking about the AI security product here and we’re not talking about the AI chief of staff, it actually honestly was not that many because what happened was so a little bit of background context for that story is that actually when I sent out that email, we’d given up on the old AI chief of staff idea and we didn’t know what were going to do next. And I had 150 investor calls lined up for the following week, right. And I had no business to pitch these investors. I didn’t even know what my idea was at this point. But because were in YC, investors had reached out and wed lined up these meetings, but just before demo day. And so we had 150 calls coming up and so were like, okay, well we sure as hell better choose something, right? 

Ravin Thambapillai
And so the first toys that we chose, I cant even remember what it was, but it was sort of AI security flavored, but it was different in some particular way. I cant honestly cant even remember what the difference was anymore. But we sent probably 100 emails for that product, like the day after we decided that the old product wasn’t going to work. And none of those got any responses. So were like, okay, scrap that. We need a different idea. And we’re on the clock here. We’ve got 14 days before the investor meeting start, so we’re down to 13 days now. And that’s when we sent the emails. And we sent probably about, I think, 50 of the AI security product emails the following day. 

Ravin Thambapillai
And two of them got responses like one from Checker and one from this big pharmaceutical company that were also working with. Now, it took a little bit longer to close them, but we did get there, and that was you send 100 emails and get zero responses for one idea, and the next day you send 50 and get two. All of a sudden the sort of part of your brain is a Founder thats like, wait, maybe it really is kind of lighting up. And so took us about 50 emails, probably sent about half of those. Jack sent half of those. But that was all it took, actually. 

Brett
This show is brought to you by Front Lines Media podcast production studio that helps B2B founders launch, manage, and grow their own podcast. Now, if you’re a Founder, you may be thinking, I don’t have time to host a podcast. I’ve got a company to build. Well, that’s exactly what we built our service to do. You show up and host, and we handle literally everything else. To set up a call to discuss launching your own podcast, visit frontlines.io podcast. Now back today’s episode. Just so we can have some context around the scale you’re operating at today, are there any numbers that you can share? 

Ravin Thambapillai
Yeah. Okay, so, I mean, we’re growing at like 20% to 30% per month. We sign up a new household name business every two months or three months. I would say we process over a million LLM queries every month. It’s pretty insane to have gone from like, crazy random idea to something that household names that I grew up using their products as a kid are now using. Credel. The team is still super small, which has been great, honestly. It’s still a five person team. We desperately need to scale the team, but we also, like, are super keen to keep the hiring bar extremely high. And so we’ve just been a little bit slow on that. But I will say if there are any great software engineers or product managers or product marketing managers out there listening to this, please hit me up. 

Ravin Thambapillai
I can’t wait to speak to you. 

Brett
What do you attribute to that growth? I think any Founder that’s listening in, wants to be able to answer that question in that same way. But for most founders, it’s hard. They’re not growing that fast. What have you gotten right and what are you doing from a marketing perspective to rise above all the noise that’s out there in the market? 

Ravin Thambapillai
That’s a great question. I think the question of what we’re doing right is sort of a funny one, right. Because from my perspective, pretty much everything we’re doing right today is something that at some point in the past, were doing wrong. Right. And so maybe the sort of takeaway from that is that actually the most important thing to do right is to actually sort of be reflective, embrace the reality of what’s not working, and then act decisively to course correct against that when something is not going the way that you want it to go. And I think that’s an extremely important part of getting to something that will work. And, you know, at the end of the day, I’m not going to deny that luck played a big part in our role. Who knows? 

Ravin Thambapillai
Maybe if Michael had just been out on holiday that week, he may not have responded to my email. I’m not sure. But there’s obviously a component of luck that’s involved. And so if you act decisively on the things that are not working and keep iterating really fast and course correcting really fast, that’s probably, like the best way to maximize your chances. And then there was a second component to your question, which I think one of the things that is really important when you’re in this kind of crazy AI hype wave that we found ourselves in is actually not getting distracted by, oh, look, this new model just got released and, oh, look, this company is saying that they’ve got everything that you could ever imagine to be fully automated. 

Ravin Thambapillai
You’ve got to ignore all of that stuff and just stay really focused on, okay, you have real customers. Those customers actually have pain points. They’ve got jobs that they’re trying to do, and those jobs are difficult in a variety of ways. What are the things that make that really hard? And how do I, like, take away the difficulty from a customer? And if you keep doing that every day, your customers lives are getting easier and easier because of the features that you’re delivering, then there’s, like, a pretty good chance that you’re doing something right. And honestly, if you stay focused on that stuff, you don’t even notice, like, what’s happening in the, like, broader universe of the noise, as you say on Twitter. 

Ravin Thambapillai
One thing I’ll just quickly say on this is that maybe for us, in some ways, this is a little bit easier because being focused on the security component and the compliance component. Security and compliance is always the most boring part of any technological wave. And so maybe we’re less liable to get distracted. But I think that the principle is still the same. Like, if you just stay ruthlessly focused on talking to actual customers, understanding the actual pain points and solving those actual pain points. Everything else, like, will fall into place. 

Brett
Is most of this growth then coming from inbound or is it coming from outbound? 

Ravin Thambapillai
Yeah, great question. Well, the funny thing is, after that round of 50 emails that we did, we actually did not do any outbound for the longest time until literally the start of this year. So pretty much an entire years worth of almost an entire year’s worth of time and everything just came from content that were putting online. Because we would work with these customers, we’d solve a problem, we’d write a case study up about it, and then the head of platform engineering at some 10,000 person public financial services company would reach out being like, oh, I read this thing that you wrote. It was pretty interesting, can we chat? And then obviously my job will be to find a way to turn can we chat into, okay, can you pay me hundreds of thousands of dollars, please? 

Ravin Thambapillai
That was a lot of work in and of itself. So I had so much work to do just keeping my head afloat on getting these leads to convert and getting the product to actually work. I just didn’t have any time for outreach, to be honest. And it was only like at the beginning of this year, we could sort of put our head above the parapet a little bit and be like, oh, wow, we have a product that works. We have all these customers. Let’s go and find the next generation of customers that we actually started to do outreach. So for us at least inbound was like definitely the primary driver of our growth and hopefully it stays that way. I definitely expect it to. 

Ravin Thambapillai
One of the nice things about once you have a solid customer base, you start to understand who your ICP is, who the buyer is, what does the buying cycle look like? And then when you have that, it’s really easy to do targeted outbound because you know exactly who’s going to be buying you and you can just hit those profiles. So that helped a lot with like bootstrapping our outbound process. 

Brett
Who is the ICP and what does that buying journey typically look like? 

Ravin Thambapillai
Yeah, great question. So the primary ICP for us is like the platform engineering lead. You look at these big 1000 to 10,000 person technology companies, they typically have like some kind of platform team or maybe it’s an AI team, and the engineering lead of that team is trying to build a bunch of applications. They have some applications probably already in proof of concept stage and they’re trying to roll it out to the entire and the ciso@the.org is like, wait, hang on a second. Have you thought about all of these security problems? And then the engineering lead’s like, well, I did kind of think about them, but I didn’t really want to deal with them, to be honest with you. 

Ravin Thambapillai
And that’s when they start doing a little bit of research and find us and find our content, and then they’ll typically reach out to us. Very often there’ll be like a sort of pilot period, maybe it’s one month, maybe it’s two months, where they’ll sort of get their hands on the product. And for our customers, like, once they’ve got their hands on the product, they pretty much always end realizing that it just makes a lot more sense to buy this rather than build it. But yeah, there are some customers, some companies that just really love building their own software for a variety of reasons. And sometimes that makes sense. And so that’s kind of typically the decision that they’ll have to make is like, do we buy Credal, or do we try and build some in house, like custom version of it internally? 

Brett
What types of terms are they searching for? Because I have to imagine it’s slightly different for every team and every person just because this is such an emerging space. 

Ravin Thambapillai
Yeah, that’s totally right. So one of the things that’s super interesting for us is we are this general purpose platform. We’re not trying to be a particular point solution, but our users are trying to solve a very particular problem. To take the example of transferwise, there’s a big financial services company, they were trying to solve, like, a Kyc AML problem, right? And so they would be searching, like, how do I protect customer data when using llms to do Kyc? Right. And so for us, like, anticipating exactly what the right search term to, like, go for was actually pretty difficult. So what we try and do is we try and put out, like, many different case studies across all of our customers about all of the use cases that they’re solving and the role that security plays in actually getting those use cases into production. 

Ravin Thambapillai
And that is we have this AI security guides sort of content series on our website, and we have a lot of blog posts about the specific use cases that customers have solved with us. And I think the combination of those two things has been really helpful because people read it and they’re like, oh, okay, I get that this is a general purpose platform, but I also get that it can actually be used to solve my specific problem. And that’s kind of really when we’re in the sweet spot. 

Brett
How do you think about your market category? Is this a category that’s already being talked about, or is this a category that you have to really define and eventually own? 

Ravin Thambapillai
Yeah, it’s a great question. I think so. From my perspective, it’s completely a new category. Right. Like, what I actually believe we are building is very different, I think, to people in general think, like, ought to be built, which is, from my perspective, what we’re really building is the safe and access controlled data environment for your AI employees. Because we believe that five years from now, every company is going to have hundreds of these AI employees running around doing work. They’re going to need to access company data, and they’re going to need to access that company data in a way that’s controlled and secure. Right. But if I go out and say to a customer, hey, this is what you need, they’re like, but I don’t have hundreds of AI employees. Right. What are you talking about? 

Ravin Thambapillai
And so when we explain the category, we do it in this very high level sense that people pretty much agree now that there is this sort of AI security category. But most companies have not actually figured out what that really means. Like, there’s all these different ways that you can talk about security. You can talk about prompt injections, you can talk about hallucinations. There’s this entire universe of AI security that people don’t really understand, like, what is real in that environment. And so we sort of think of explaining the category as AI security, but in practice, like, what we’re really doing is defining what AI security really is. And what that is giving AI applications and employees a safe and secure way to access company data. 

Ravin Thambapillai
And so that’s like a gradual journey we have to take our customers on, and eventually they realize, but typically in the first meeting, they don’t really know what AI security actually means. 

Brett
What I tend to see is when there’s an emerging category like this, a lot of vendors end up essentially just battling to say, okay, here’s our definition. And then they try to force that into the market and evangelize that vision and get the market to embrace and accept their vision. What are you doing to ensure that the market buys into your definition of AI security and not another vendors? 

Ravin Thambapillai
Yeah, great question. So from my perspective, right, the thing that actually matters is not like, are you making AI secure? That’s obviously a nice instrumental goal, but the actual goal is to make sure that these enterprises can actually deploy AI against extremely valuable use cases and then ultimately trust that AI is being deployed in a secure way. And so what we’ve really focused on is not so much like, here’s our definition of what AI security is. Market go out and accept that. What we’re really focusing on is asking our customers, hey, what are the actual things that you’re trying to do? One of the problems, what are the security concerns that you actually have in real life? And which of those security concerns are theoretical and which of them are the ones that you’re actually completely worried about? 

Ravin Thambapillai
When you’re building an internal AI application, there’s some sort of theoretical worry that one of your employees might go rogue and try and inject some nonsense into a pump. But typically thats something that you can risk accept because your employees are actually pretty trustworthy and you’ve done all these background checks on them and you give them a high degree of autonomy of the company anyway. But what you actually cant risk accept is the idea that an honest employee might log into the system and ask a question like, whats the company’s strategy? And the AI accidentally retrieves an email from the CEO’s inbox thats like, hey, were going to fire the entire engineering tomorrow, right? And that’s not something that you can just risk accept. You actually have to freakin deal with that potential issue. 

Ravin Thambapillai
And so that’s what we try and do, is we try and just ask our customers, like, which of these security problems, at the end of the day, are you just going to risk accept and which of the ones are you not going to risk accept and you actually need to solve? And then when we take our customers on that journey, it tends to emerge that the core problems that people really feel like they need to solve. How do I manage the permissions of my data? And how do I make sure that my use of AI is compliant with the law? How do I make sure that it’s auditable, that it’s fair, that it’s unbiased? Then I can go to a regulator and say my use of AI is defensible. 

Ravin Thambapillai
Those are the actual problems that no matter how you want to explain to the market, you should care about this. These are the problems that unlock value at the enterprise. And that’s what we’re really focused on is like, what do you need to do to unlock value? 

Brett
We’re already about two months into 2024, so I’m sure you’ve already finished up your 2024 planning in that 2024 plan. What are those top priorities? And then what’s keeping you up at night? When it comes to thinking about making those priorities happen and executing on those priorities. 

Ravin Thambapillai
Yeah, another great question. So our mission is really helping enterprises make AI safe and useful in the industries that matter the most to human welfare. So our 2024 plan is really all about continuing to realize that. So today, I think our product does just an incredible job of operationalizing AI on unstructured data, right? Documents, PDF’s, images, videos, that sort of thing. But obviously a massive amount of company data is actually like the structured stuff, your data warehouse, your production databases, the sensors on your drones, all that stuff. And so we’ve really only just started delivering against that vision recently, but we’ve already seen huge demand to be able to do the sorts of things that these models can do, rather than it just being on like simple text for it to be on a billion row SQL table or something. 

Ravin Thambapillai
And yeah, there’s a huge amount of product work to do there. And then I think beyond that, I do actually still eventually expect the market to mature and for that to be like hundreds of these AI vendors trying to sell AI applications. And those applications will need access to various bits and bobs of your data. In this conversation right now that we’re having, there’ll be an AI that briefs you for this podcast, and that AI will need to talk to the AI that briefs me for this podcast, right? And yours will be sharing themes most relevant to the audience, and mine will be sharing the data that’s like relevant. And they’ll be going back and forth to try and figure out what are the interesting topics for discussion, for credible. And then obviously that will be slightly different for the next guest. Right? 

Ravin Thambapillai
And that exchange of information will need to be governed because I want the AI to help brief you, but I also don’t necessarily want it to accidentally reveal that I’m trying to get Ilia fired or that I’m in a dispute with Satya Nadella or whatever. And obviously that’s a joke. But the point is that governing that sort of interaction in a way that’s deterministic and not actually left up to chance is going to be really important. And I think Credo is honestly the only business that can actually do that. So thats our big 2024 vision. Now, in terms of what keeps me up at night, 100% of that is the talent component. So at Credo, were just obsessed with having insanely high talent and insanely high caliber talent. 

Ravin Thambapillai
I think part of our hiring process is after the interview process, when you’ve wowed all of our employees during our reference process, well actually ask your references. Hey, is this person actually the best person you’ve ever worked with? Maybe one of the best two or three people you’ve ever worked with. Right. And if we don’t get the references to say, like, yes, this is one of the best people I’ve ever worked with, then we’re just not able to make a hire because that’s what we expect amazing people want to work with. And so I, as a Founder, have chosen to hold this insanely high bar. It’s pretty scary to me that might be a mistake, right. Because obviously, the higher you hold the bar, the narrower the talent pool that you can hire from is. 

Ravin Thambapillai
Now, I will say that we’ve been lucky so far in being able to hire, like, an absolutely exceptional team, but we just have so much work to do, and scaling the team whilst preserving that bar is a daunting task. And it’s definitely the thing that I think is most likely to be a disruptive force for our ability to execute my vision. So, again, going to just repeat the call to action that I had a while ago, which is, if you guys there any amazing software engineers or product marketing people out there listening to this, please, I really want to hear from you. I really want to build a future with you. So, yeah, get in touch. 

Brett
Final question for you before we wrap up here. Let’s zoom out three to five years into the future. What’s the big picture vision? 

Ravin Thambapillai
Yeah, great. Well, kind of what I said just a second ago, we’re making enterprises able to actually deploy AI in a way that’s both safe and useful in the industries that matter the most to human welfare. And what that actually means, right, is the access controlled data lake for your AI employees. So all of the company’s data, whether it’s structured, unstructured images, videos, drone sensors, everything, all together in one place in a way that AI can actually understand that data and can also understand who should have access to what parts of it. And that’s the core of what we’re building. 

Ravin Thambapillai
We believe that the long term story for AI will be, it won’t be like one workplace assistant, it will be like hundreds, maybe thousands of little workplace assistants that will have specific jobs to do and will need to access, like, different parts of your data. And the AI’s will need to communicate across, not just within a company, but across companies. Right? Google will need to talk to the AI at Google’s customers and vice versa. And so I think every enterprise is going to need a good way to govern that sort of behavior of these AI’s. But also to control and secure that behavior. And fundamentally, I think what that does is it is actually the key unlock for enabling this incredible technology to actually be valuable in the industries that are so important that the government cares about what you’re doing. 

Ravin Thambapillai
So healthcare, education, financial services, housing, these are the really critical industries. And I think these are the industries that are going to care the most about the security and compliance component of how they deploy AI. So we want to make it safe, we want to make it useful, and we want to create a world in which AI is actually realizing its potential for human welfare and isn’t just like some toy. So thats what were trying to do. 

Brett
Amazing. I love the vision and I’ve really loved this conversation. I know its going to be a big hit with the audience. We are up on time, so we’re going to have to wrap before we do. If there’s any founders that are listening in or talent that’s listening in, where should they go? 

Ravin Thambapillai
So you can find us on LinkedIn or at our website, Credal AI, that’s Credal dot AI. And then you can also just reach out to me straight up at RavenCredal AI if you have questions or you’re curious about working at Credal. Amazing. 

Brett
Raven, thank you so much for taking the time. Really appreciate it. 

Ravin Thambapillai
Thank you very much. Thanks, Brett. 

Brett
Thank all right, you keep in touch. This episode of Category Visionaries is brought to you by Front Lines Media, Silicon Valley’s leading podcast production studio. If you’re a B2B Founder looking for help launching and growing your own podcast, visit frontlines.io podcast. And for the latest episode, search for Category Visionaries on your podcast platform of choice. Thanks for listening, and we’ll catch you on the next episode.