Senior Software Engineer, AI-Driven Systems

### **About Us** Thimble was an early pioneer in insurtech. We built a digital-first way for small businesses to buy coverage by the job, month or year, and have since delivered more than 200,000 policies across the United States. We started direct-to-consumer as an MGA, but today, we’re focused on growing through brokers and distribution partners, bringing the same simple, fast experience to the agents who serve small businesses. Thimble is still evolving and is at the front of AI-native distribution with MCP technology paired with API distribution. In 2023, Thimble was acquired by Arch Insurance, part of Arch Capital Group Ltd. (Nasdaq: ACGL), an S&P 500 company providing insurance, reinsurance and mortgage insurance worldwide. That gives us something rare: the speed, ownership and culture of a startup, with the balance sheet, stability and long-term commitment of one of the industry’s strongest insurers. For more information, visit [](http://www.thimble.com/). ### **Our Values** Our unique culture is built around our six core values: **Think Wild, Act Tame** Be relentlessly creative. Question everything. Get to know the rules so well, you can use them in new ways. Be the industry’s compliant radicals. **Green Means GO** Act. Move forward. Learn. You can only make a decision based on the best information you have – so make it. **Own Every Detail** Be ruthlessly precise. Disciplined. Sweat the small stuff. **Simplify,** ~~**Simplify, Simplify**~~ Get to the point. Make it crystal clear. Cut everything else. It’s our job to make insurance easy – however hard that is for us. **Every Moment Counts** Show up early and ready. Deliver on time or before. Make the most of your time, so others can make the most of theirs. **Stay Human** Treat every person, and their ideas, with respect. Cherish diversity. ### About the Role Thimble is investing in the next generation of its platform, and this engineer helps lead that work. Our backend is built on Node.js microservices and event-driven architecture on AWS, and we are evolving it toward durable, explicitly orchestrated workflows, with AI-native tooling at the center of how we build. This is a hands-on role: you’ll write, review, and ship production code every day, with AI agents as your primary build partner. You’ll also shape technical strategy and set the patterns the rest of the team builds on. You’ll work in Node.js and TypeScript every day, so you need to be productive in them from the start. ### What You’ll Do - Design and build the durable workflows and services that will replace our current event-driven flows, defining service boundaries, state ownership, idempotency, and retry semantics - Use of agentic coding tools on hard problems: setting up context, breaking down problems, and verifying output - Write and ship production code in Node.js and TypeScript every day, using agentic coding tools (Claude Code and similar) as your primary way of working - Define the reference patterns (service templates, shared libraries, code generators) that every rebuilt service follows - Plan and execute the migration in increments that keep production running, with tested rollbacks - Build and maintain the tooling around our AI workflow: MCP servers, evaluation scripts, and automated checks that hold AI-generated code to the same bar as hand-written code - Instrument services with metrics, alerts, dashboards, and tracing, and own production incidents end to end - Make targeted changes in our React and Angular applications when backend work requires them - Partner with product, growth, and other engineers, and contribute to code reviews and architecture discussions ### What We’re Looking For - Strong backend engineering with production experience in Node.js and TypeScript - Hands-on experience building and operating distributed or event-driven systems, and a working understanding of how they fail - Experience using AI tools to drive a large code migration at scale (framework upgrade, service rewrite, language port) - Curiosity about agent harnesses, MCP, and A2A protocols; production deployment of agents is not required - Experience contributing to a migration or re-platform on a live system - Comfort reading unfamiliar code and data models and spotting problems in them - Enough frontend literacy to trace a problem from the UI down to the service layer - Clear written communication and comfort working in a remote, async-friendly environment ### Bonus Points - Hands-on experience with a durable execution or workflow orchestration engine, in any language - You’ve built or contributed to an MCP server, an agent harness, an evaluation suite, or a multi-agent system, whether as a side project, open source, or an internal tool - Experience in InsurTech, FinTech, or another regulated industry in a similar environment to our tech stack