Pace raises $46M Series B to automate insurance back-office operations with AI agents

Have a good one.

Thank you. Goodbye. Uh,

has to be has to be one of the most powerful names in

Yes. Mount Castle.

Mount Castle. Mount Castle. Like

arguably the most regal name in town.

It's a fantastic name and he's been fantastically successful. Um, but we have Jamie from Pace here in the waiting room. Sorry for keeping you waiting.

Welcome.

Not at all. It is great to be back. Thank you.

Thank you so much for hopping on.

You've been incredibly busy.

Yes. Give us the news. Give us an update. What's going on?

Yeah. Well, we are super excited today to be announcing our $46 million series B

collab by driving with participation emergence and proven. Yeah, super pumped on news.

So, uh what has been the key unlock? I mean, insurance operations. Uh obviously there's a lot of intuitive applications for AI and insurance. I think people understand how this product could be valuable, but uh what is the shape of the gotom market motion? Is this something where you get one big insurance company and like that's 50% of the market and you're good to go? Are you working with startup insurance providers? We've talked to so many different people that are figuring out how to ensure data centers and all sorts of different new things and underwriting and new pools of capital, but where's the been where's been the biggest growth source for you?

Absolutely. So, PACE is the AI operations partner for the world's leading insurers. We primarily are focused on the world's largest insurers. So, large carriers, large brokers like Credential, WW, Convex, um that are our customers today. And we're really helping them to automate a lot of their back office operations and help ensure more of the world's risk.

Uh what are the biggest challenges right now? You know, we've talked with uh people uh from all different types of companies selling agents into the enterprise. Uh there's uh plenty of workflows where if you're getting things right 90% of the time it's like not even close to good enough, right? But then there's others where like that's a good kind of like first pass and just makes the existing back office team more efficient. So like what's working? Where do you want the models to get better? How are you kind of taking that into your own hands? Uh I don't know if you guys are doing your own you know um post training and things like that uh or or you know leveraging open source models but um but yeah what is the current state of things?

Yeah absolutely. So I think that's definitely very true in our regulated market where you know many of our customers have 99.9% plus accuracy SLAs's and we are hitting those day in day out and they have to cuz that's uh you know that's the um the way that these get into production for our largest customers. Um, you know, for us, some of the things I'm like really proud of as a statistic for our company is that we've had a 100% win rate from pilot to production. And a lot of that is because we, you know, it's the product and then our forward deployed engineering team that are working with our largest customers to help them, you know, um, successfully deploy these agents into production. And so I think the the a lot of what we're seeing in terms of you know being able to deploy and hit those sort of accuracy levels and really see our customers getting you know the um the automation rates and the ROI they want is really combination of the insurance expertise that we have as well as the AI expertise and kind of bringing that together. What's that what that has looked like for our customers is we have customers that are you know require that 99% plus accuracy. We have customers who requiring hundreds of thousands of tasks to be automated every day in the background and we're doing that fully autonomously with agents. You know, we have, for example, Palomar, one of our customers has said 90% um of their uh tasks are being completed from intake to outcome um successfully by pace agents fully autonomously. And so, you know, we we've been super fortunate to see that in production. In terms of, you know, what has been a massive accelerant to our business in doing that, certainly it's been the models getting better. we have very much lent into these agent operating procedures. So sort of natural language instructions that the agents use to complete these tasks long running. And as the models get better, our product has been getting um even stronger. And I think one particular area we've seen that is in computer use. Um a lot of our customers had challenges in the past deploying products because they needed to integrate with various APIs either were too expensive or you know just prohibitive to build out or didn't exist. and computer use models, they've really gone from something like 30% accuracy on our evals to like 95% plus. And that really allows us to get uh these models into production. Um and particularly work around cases like desktop applications or legacy web apps or even like we've seen like green screen CLIs uh you know and being able to have the agents um get live and do this work fully end to end. So those are a few of the unlocks for us recently.

That's great. Uh another question, there's been uh some new insurance startups uh that have been growing very quickly recently and there's been some debate around uh h ultra high growth when you're in a when you're when you're basically uh in the risk business can be uh risky. Uh how do you know the companies that you guys work with think about growth? or are they really just trying to speed up the back offs? Uh speed up the back office, speed up just like you know increase efficiency so when a new customer comes in they they get them you know onboarded as soon as possible. Uh but like what is their general philosophy around growth given that these are companies that are at scale? I imagine they're not coming to you saying we want to be able to grow three times faster by doing this. It's more creating efficiency on the back end. But what do you think?

Yeah. So, you know, for our customers, these are, you know, in many cases like hundred-year-old plus businesses that are thinking about what does the next, you know, decades and centuries look like as they want to, you know, serve their customers better and better. Um, and I think that one thing that's like really top of mind for the business is the protection gap. So, last year 60% of the world's losses were uninsured. So that means, you know, a house after a hurricane that's flooded, you might be uninsured or, you know, life insurance. There being just a massive protection gap between what makes sense, you know, for a family having life insurance versus what they have today. And I think a lot of our um a lot of our customers are thinking about that's their kind of responsibility is how do we shrink that protection gap? How do we take the $9 trillion dollars of risk today that is uninsured that should be insured and enable that to to happen? And so, you know, for us, what I think we've been most excited to see is when you take the a the operations component with AI native operations, you can truly change the economics. Um, like what we are seeing is two orders of magnitude less spend than would be been needed before with agents. And what that enables is our customers to launch new products which they have um to be able to deliver an experience to a 10person company that they that is similar to the experience that they would have previously only been able to provide to a 10,000 person company for a claim to be paid consistently on time every time. And so that's like really what our customers are pushing towards which is how can we close that you know protection gap enable sort of the 60% of these losses that are uninsured to be insured. And that's truly, I think, what is so exciting about this opportunity is like this is the moment that we can really unlock the um the the sort of operations that makes that possible.

Very cool.

Fantastic. Well, uh I have a bunch more questions, but we'll have to do it the next time you're on the show because we've been running late.