Integral raises $18M Series A to sanitize proprietary data so AI builders can access sensitive datasets

for going public.

Hey, if you need to raise capital, if you want to raise the most capital, you got to go to the New York Stock Exchange. That's the endorsement that we're proud to give. Our next guest is in the waiting room. Uh Shub Shinha from Integral is the co-founder and CEO. Welcome to the show. How you doing?

I'm doing okay. Thanks. Thanks for having me.

Just okay. Just okay.

On this big day,

I'm uh I'm riding the highs and uh and the lows. and getting getting back up right now. So excited.

That's that that's pretty normal for the entrepreneurial journey, but uh take us through your entrepreneurial journey. How did you get here? What are you building? What's the news today?

Yeah. Yeah, definitely. Well, I'm excited. Uh today we actually just announced an $18 million series A for uh for my company, Integral. And so the gone, it was a good warm-up hit. Had to get the real hit in there. We like to warm up the gong on this show.

Double hit. I don't think we've done that before.

There's a warm-up hit and then the real hit.

Yeah. Yeah, that's what

that was special for integral.

So tell us about integral.

Yeah. So what integral does is we sanitize proprietary real world data sets such that AI builders can get very bespoke, very sensitive data sets.

Um but that data holders can also make sure that privacy and compliance is adhered to. And so this things this looks like medical records, financial transactions, etc. You know, a lot of this contains the real world human behavior patterns that people like you and me have. And there's a ton of individuals and enterprises who are now monetizing it since the AI companies and the AI builders want it.

That being said, there the the builders want the signal. They don't want the secret values. They don't want the the the proprietary secrets, so to speak. And so integral sits in the middle where we ensure that through privacy engineering both with our privacy teams and our engineering teams we can retain the utility and the signal while also ensuring that privacy and regulatory and contractual.

Yeah. So how about this for an example? There's there's been uh you know there's been examples of of you know a company buying a company just for the data around how that company was operating like just for the slack. But then you can imagine as Slack there's so many things

that uh so much information in Slack that again you talked about as signal but not necessarily information that should be available for even people let's say working at the labs to see right that's like

PII etc. That's right. And and folks at the labs don't even necessarily want to see that, right? They want the context of it all. And so, it's a real win-win for data holders who want privacy and trust and also revenue. And then the lab uh who or or the AI builder who wants to ensure that they're not stepping into anything they shouldn't be stepping into, but they get that value.

What are some types of companies that have valuable data that that don't know that it's valuable and it would be something that they could monetize if they were, let's say, working with you?

For sure. And and that's that's one of the bets that we're making that there's this new proprietary data economy of individuals and enterprises. And so we've seen kind of all across the board just given the the kind of newness of the economy, the types of data sets that are coming in and the bespoke demand. So we've seen uh for example like mid-market hospitals that see compressing revenues but have a very very specific treasure trove of data uh because they focus on a rare disease or a specific procedure and and that's not available anywhere else. fitness apps. Um, some people, you know, they log their entire lives into fitness apps. And these apps have free or paid users.

They're only making money off the paid users, but they can then those apps can then monetize what is that entire real world signal and really continue to make that app free and introduce a ton of value. Um, and it all becomes circular because as AI gets better, a lot of these same companies will actually use it. Sure.

So, it's a real win-win there.

What's the state-of-the-art in uh sanitization, maintaining privacy? We talked to Ed and Tai from the National Design Studio on Monday. They uh shipped a something like less than 15 meg like small language model for sanitizing uh documents in the browser runs very efficiently. Uh are you just throwing open source models of this? Do you need frontier models for data sanitization? And then is there some worry where it flows back if you're using some closed source model and the frontier lab takes the data that you sent them and and stores that improperly?

For sure. For sure. And and there's a variety of solutions out there mixed with call it human services and entire teams of PhDs looking at it.

Might just be like reaxes, right? Right. Like if it's a phone number, identify it, just turn it into X or number number sign or something like that. Right.

That's right. That's right. So there's a ton of solutions that kind of do that first layer. I think where Integral really specializes is, you know, we got our start healthcare and healthcare has all proprietary data. You know, your doctor should not be putting your internet on or your data on the internet. And to and as far as we can tell, they're not

I told my doctor, go ahead.

Yeah. because it would be embarrassing for everyone else if like your your lean muscle mass and your body fat percentage got leaked out there. Everyone would

if your if your test levels leaked on the internet, it would it would be it'd be crazy. You'd be getting accused of doping.

Um

uh are you when did when did you make the switch from AirPods to to wired?

I'm also a wired a wired headphone guy, so I'm I'm I'm curious. I hit number five on air replacement and then I went wired and it turns out so I'm I'm in New York and it turns out it's a fashion statement as well. So I'm I'm kind of rocking both sides.

I think it is uh entirely superior and and that is even before the fact you can get like a bunch of pairs of wired headphones, have them everywhere

and uh for the cost of just one pair of AirPods. So

maybe maybe

we're we're ahead. Uh congratulations on the round. uh business makes a lot of sense and excited to see where you go go with it.