Senior Engineer

Build the agent systems regulated businesses run on. Engineers first We hire strong engineers who build with AI tools every day. The bar is production rigor, shown however you have earned it: shipped products, open source, or research. AI-native by default Our engineers build with our own products, systems and tools, including a team of agents that handles the repetitive layers, so your judgment goes where it matters. Embedded and senior Much of our work is embedded, building alongside the teams we work with as a founding-level engineer or product lead. You own the outcome and the relationship, all the way through. What you would work on - Agent harnesses: the loop that calls the model, routes its tool calls, keeps the context straight, and recovers when a step fails. - Systems inside someone else's estate, where you would debug things you did not build and earn the trust of their engineers. - Work where evidence is the constraint: approval gates, audit trails, least-privilege access, safe failure modes. Engagement: contract, embedded. Level: senior and founding-level. Stack: TypeScript, Python, Rust, agents. Nice to have - Experience in or around a regulated setting: financial services, health, the public sector. - You have built evaluation harnesses, or the tracing that makes agent behavior legible after the fact. - Open source work, where anyone can read how you build.