Applied AI Engineer

About QuoteWell We started as a vertical SaaS company in insurance. We grew into operating our own wholesale brokerage and a programs group. Through that, we've lived the operational reality of applying AI in insurance — what works in production, what breaks, what compounds. We've just closed fresh funding to scale this operating model to retail insurance agencies across the country. Our CEO, Joey Bouchard, built and ran applied AI teams at Palantir. We bring that technical rigor together with years of hands-on insurance operations experience to help agencies deploy AI that works in the real world. (Why insurance is worth this attention. https://medium.com/@jbouchard_6468/why-insurance-is-awesome-e5a00daa752b) Why this role exists Brokerage execution runs on heroics. Service quality depends on who owns the account, how overloaded they are, and what they happen to remember at the right moment. A submission gets marketed broadly, or it doesn't. A renewal gets the same rigor as new business, or it doesn't. A missing detail gets chased, or it stalls. We don't think that's inevitable. Brokerage execution can be made systematic: software, automation, and AI carrying disciplined follow-through alongside human experts, inside clear bounds, in a regulated environment. The hard part isn't the model call. It's governability — making AI accountable enough to trust with real deal flow. Who is the AI acting for? What is it allowed to do? What happens when it's wrong? In a regulated industry, those aren't side questions. They're the engineering problem. We've been answering them inside our own wholesale brokerage for years. That's where we test whether the model actually produces more reliable, scalable, and consistent service. If it doesn't work for us, it doesn't deserve to be packaged for the market. Now we're packaging it. The role Applied AI Engineers sit at the point where Terminal, our platform, meets the messy reality of a customer's agency. You own the outcome at the customer. You also own the signal back to the rest of the company. A typical month: you're embedded at a retail agency, working alongside their operators. You're identifying the workflows where heroics break down — submissions, renewals, follow-up, the work that today depends on someone's memory — and you're building governed AI execution into those workflows on Terminal. You're calibrating where humans stay (judgment, relationships, accountability) and where the system carries the standard (follow-through, coordination, attention). The hardest and most interesting part: making it trustworthy. How AI decisions get reviewed, attributed, and defended. That work is what determines whether any of this can actually live in production. Then you bring it all back. Engineering, Product, and Design need to know what generalizes and what doesn't, so the platform compounds across every customer instead of fragmenting into one-offs. You're the conduit. What you'll do - Embed at a retail agency and own customer-side execution end to end - Build governed AI workflows on Terminal — integrating with AMS, carrier portals, email, and the systems an agency actually runs on - Identify which parts of an agency's operation can be made system-governed without losing the judgment that has to stay with people - Design the review, attribution, and override patterns that let a regulated business trust the system - Carry field signal back to Engineering, Product, and Design so the platform scales rather than splinters - Represent QuoteWell with customers — scope, troubleshoot, and reason about problems out loud, in their conference room - Travel to customer sites ~25% of the time What we're looking for - 1-5 years shipping production software (Python and/or TypeScript a plus) - Strong engineering foundation — CS degree typical, talented dropouts welcome - Hands-on experience with modern AI application patterns, or the aptitude and hunger to get there fast - Someone we can trust in front of a custo