00:06:15 · Angela McNea, Ruby Justice Thelot
[00:06:15] S0: So, what is the operability of the system that's created? [00:06:18] S1: Yeah, when did you start to see that this was a problem? I think this was like, you guys raised in 2025? [00:06:24] S0: We raised our A in 2020 Yeah, yeah, So Maya and I left in spring of twenty twenty three. [00:06:34] S1: Early in AI days Our [00:06:38] S0: thesis from the beginning was the glue, the harness, the control, the governance, all of these pieces were going to be important. And it's been fun to sort of watch your hypotheses play out live. And we've always been sort of pragmatists in how we build. We're like, okay, this is a challenge for people now at the best of times and this is just going to be compounded when builder ecosystems grow, model providers expand and grow, you're going have frontier models and sovereign models and local models, it's just going to compound. [00:07:15] S1: Very prescient bet that you guys were making. So I'm very curious about what inspired you to write Every Good Decision is an ensemble. Because we see now, we talk a lot about model routing and how it's very important to have something that can choose what level of model based off of some sort of cost trade off. And then you sort of make the... You zoom out and make the argument that everything has always been this sort of ensemble and not just like one model or entity making decisions. Whether it's how the judicial system works or whatnot. I'm curious, how does that connect to Thread AI?
00:07:54 · Angela McNea, Ruby Justice Thelot
[00:07:54] S0: Yeah. So, again, was seeing this explosion in conversation around routing. And there are real important conversations to be had, we're seeing it with our enterprise customers of like when is it too expensive to use this kind of model or the latency is too high or there are real tradeoffs around the specific nodes. The interesting part to me is that the actual hard part, the technology behind a router, don't think is as interesting as the The hard part is the application around how the business person or the engineering teams decide what is the logic of what models do we pick, what are the things that we're trying to optimize. And then you have to predefine all of the decisions upfront, saying like, okay, well, we've decided that this is what a hard problem is. This is the shape of a hard problem is. We've decided what is expensive. How do you define expensive? Is it expensive in time, in people's time, in money? So you have to predefine all of these like, business and, you know, esoteric and abstract things ahead of time before you even get to, cool, now we just need to, like, something to help me route between these systems. And so, we define, we build our platform so you can build these kinds of routers natively in our product. But we spent a lot of time on those decision gates explicitly and those logic gates because we were seeing like in the real world, those were the actual barriers to putting something in production. [00:09:34] S1: Yeah, I see. I want to
00:09:34 · Ruby Justice Thelot, S2, Angela McNea
[00:09:34] S1: Yeah, I see. I want to Ad[00:09:36] S2: shout out our sponsor, Lovable. You know, the app you've been meaning to build? The internal tool, side project or product you'd otherwise lose a weekend to? With Lovable, you get software you can ship today. That includes an editable code base you can inspect and change, plus two way Git GitHub sync. Lovable also handles the backend and infrastructure, including managed Postgres, auth, storage, hosting, payments, and more, so you don't have to wire it all up yourself. And through Lovable's MCP server, you can create and deploy your project directly from the agents you already use. Turn ideas into software people love with Lovable. Lovable. Dev. Now onto the episode. [00:10:14] S1: So I think, like, you know, the topic of agent failure can be very high stakes Mhmm. Specifically with anything to do with defense or anything to do with national security. [00:10:25] S0: Mhmm. [00:10:25] S1: So I'm curious about your, you know, your, like, defense threat AI and how it helps in defense use cases or defense providers?
00:10:25 · Ruby Justice Thelot, Angela McNea
[00:10:25] S1: So I'm curious about your, you know, your, like, defense threat AI and how it helps in defense use cases or defense providers? [00:10:33] S0: Yeah. Being a dual use company was very important to me and Maya from the beginning. And there's no more demanding customer than the US Department of War, Department of Defense. And our same primitives, same building blocks that we use in those environments are the same that power our enterprise work. And I could talk about sort of from the beginning, Maya and I sought out Hard Problems We Want. Our goal was to help companies and organisations and institutions, like, fix hard problems and help people. So a lot of our work... The one that we sort of talked about publicly recently is we are doing some research with Army DevCom. And a lot of that work is around manual deconfliction. So how do you help people get accurate information at the right time when you're sort of building and maintaining existing systems, control systems, and innovating new ones? So how do you help the scientists, and how do you also help the warfighters on the ground? And again, that's probably one of the most demanding environments I can think of. [00:11:48] S1: Yeah, totally. I mean, I'm just curious if you could paint a picture of the stakes for agent failure in this industry, but also others.
00:11:48 · Ruby Justice Thelot, Angela McNea
[00:11:48] S1: Yeah, totally. I mean, I'm just curious if you could paint a picture of the stakes for agent failure in this industry, but also others. [00:11:58] S0: Yeah. So again, an enterprise version of this is if you have agents that are helping you with your reconciliation, again, common one, reconciliation flows in sort of a big financial institution. Mistakes there can cost your business billions of dollars. So they carry a huge financial and potentially, probably actually a legal risk element to it. And then a of defense setting, it's like are life or death stakes. So you have extrema. [00:12:39] S1: Yeah, definitely. So I'm curious too, what's next for Threat AI? [00:12:45] S0: So we're growing quite a bit. We have a bunch of open roles around some of our public sector work in particular. So you'll be seeing more from us on that element. And then one of the things I'm pretty excited about is that sort of the investments that we've been making around our building blocks. So for companies who are selling products or building products that they're selling to their customers and want those products to be more adaptable and agentic and AI forward. We've been providing sort of building blocks for companies building products. Keep an eye on some of the stuff we're putting
00:12:45 · Angela McNea, Ruby Justice Thelot
[00:12:45] S0: So we're growing quite a bit. We have a bunch of open roles around some of our public sector work in particular. So you'll be seeing more from us on that element. And then one of the things I'm pretty excited about is that sort of the investments that we've been making around our building blocks. So for companies who are selling products or building products that they're selling to their customers and want those products to be more adaptable and agentic and AI forward. We've been providing sort of building blocks for companies building products. Keep an eye on some of the stuff we're putting [00:13:25] S1: out there. Yeah, Awesome. So, I'm curious, like, what... From from, you know, looking at these companies, if you have any takes on sort of the the ways that companies are just, like, structurally changing and maybe the role of, like, human agent interplay? Like, what, from your perspective, like, where are we going?