So what is the, the boogeyman here? Like, what are people worried about? If you're a family office professional manager and you have a suspicion or you know that people within the organization are leveraging public LLMs, so we'll use Claude or ChatGPT as maybe examples here, and onboarding sensitive information, what are the implications? What are the risks associated with that?

Yeah, I mean, look, everyone's gonna tell you that data privacy is the biggest risk associated with these types of things.

Uh, and I, I, I tend to look the, the, the model provider's interests are not aligned with your own when it comes to data privacy and tooling.

And in fact, whether it's overt, covert or straight up illegal, they will find ways to use your data.

And I'll, I'll give you a couple examples of, of each of those.

So, uh, in the overt capacity, Anthropic last year just publicly stated that they were updating their terms of service, that if you were not on a very specific account type, that they would be using your data to train their models.

And that's if you were on their enterprise account.

How should people be thinking about them from a risk and compliance perspective?

Yeah

... their new Spark models are actually pretty cool.

Um, so you have unlimited use of the top six model providers, um, OpenAI, Anthropic, Google, Perplexity, Groq, and now Meta as well, all for a, a very low cost.

So that's for the end users.

Um, for the administrators of these accounts, they have complete and total governance and control over how data is treated before it ever leaves their network boundary.

And because what-- the way that we're deployed, which is called single tenancy, which just means that we are acting as an extension of their network boundary, all of their data is contained, and they get to choose from, you know, uh, uh, as granular as how a name is treated up to how documentation, how MCP calls are made, um, how that data's treated before it ever reaches an LLM.

So their end users are able to use these tools in a interface that is extremely familiar to them, um, that feels really neat.

Okay.

Yeah.

I think it was...

There was a, uh...

The way that this whole thing was founded was, you know, my co-founder, Aaron, and I, we'd come out of FIS in financial services, and my whole job there, I was their chief innovation officer, and it was a giant fintech company, and we la- tried to-- We ran a venture builder, actually.

So we tried to launch a bunch of different companies in there, and every time I would go up against our, our chief information security officer because he'd wanna make sure that everything was tightened up and met compliance and regulatory and, and met all regulatory considerations because, of course, anything that we sold was used by people that were subject to substantial regulatory consequence if they somehow violated those.

And so we, we, you know, were looking at what was happening with AI and thought, "My goodness, here comes a whole new lane of compliance that we don't have, that no one's ever thought about.

Uh, is that the place that we wanna go solve for?" And, and actually, what we did is we teamed up with another venture studio called High Alpha, and we spent a bunch of time just digging into what customers were looking for.

Yeah.

Tell me about that MVP, how long it took to build and what sort of tools you're using to bring it to life.

Now bringing it to life is different than what we had then.

I think you'll recall this.

Anybody who's in software engineering knows there was a major turn in the capability of cloud code, call it December of last year.

Prior to that, we were pretty reticent to use coding agents, actually.

We have a lot of engineering purists on our team, and they didn't really feel like the code that the models were spitting out was actually really usable.

And they would say it would force far more time in review than it would actually save them time.

it's

Oh, I was just going to ask too, like, can you dig in a little more on what are those threats with your data with the cloud providers, right? Like, it's clearly obvious, like the threat of like your employees uploading sensitive data, but what's the rest of that exposure? What happens

when? Yeah, exactly.

Right.

Well, let's wind the clock back three years because the threat still exists.

Whether or not it's as overt as it was then is very different.

And I'll give you guys kind of three examples here.

So we'll go with the kind of overt threat, the covert threat, and then we've already talked about the nefarious threat with Google.

be a question with an answer.

You know so much about this, right? Like

you're building like such a really cool, like the large enterprises are leaning into where Mistral's headed, which is driving a lot of them around the open source modeling and everything.

How do I enable that to be? How do I know my company to own my AI? And there's a lot of value to that.

However, if you kind of tap down market, like don't think large enterprise, and I'll tell you a little bit about how we got there.

So the problem being data privacy, and then security based around observability, how do I understand what's happening? How my users are engaging with models and then administration and governance.

How do I actually enforce policy across all of these different models that are going to be deployed in my organization? So break security into those four things, data privacy, observability, governance, administration.

That's kind of the security challenge that we have.

secure?

Looking back, what experience most influenced your approach to building Liminal?

Oh, [clears throat] to building Liminal, I think it's a, it's a, uh, probably a combination of different experiences.

I, I grew up in a household that valued taking risk.

My father's an entrepreneur.

He's done it several times.

Um, and I, I think growing up and understanding that there's a, a rationale to betting on yourself, to going all on yourself, and the, this belief that, you know, "Why can't I do that?" Um, you know, "If they can do it, why can't I do it?" uh, is, is a really big part of why I started Liminal.

It's a big part of why I started my first company.

You can do anything, you know? And I'm concerned about it, you know? And, uh-