Member of Technical Staff, AI Engineer

Testudo•New York City, NY

2d•Onsite

About The Position

As a Member of Technical Staff, AI at Testudo, you'll build the core services of our underwriting and risk monitoring platforms — the systems that decide what we insure, what we charge, and what we watch after the policy is bound. You'll be among the first engineering hires outside the founding team, with end-to-end ownership of features, real product input, direct access to customers, and the mandate to move fast. The systems you own will manage billions of dollars of real financial risk capital.

Requirements

5+ years building production web software, ideally at a startup.
Fluent in Python and Node.js/TypeScript, with real front-end and design experience in React/Next.js.
Designed scalable services, built high-throughput data pipelines, or handled real-time monitoring.
Operate with ownership and autonomy; you bias toward action and unblock yourself.
Communicate well across product, risk, data science, and customer contexts.
Relish direct feedback and want to work in person, at pace.
Shipped LLM or agentic features to production, and built the evals and observability that tell you when they are wrong.

Nice To Haves

Need the architecture settled before you start building.
Want heavy process and long planning cycles over a ship-daily culture.
Would rather own one layer deeply than the whole path from data pipeline to dashboard.
Are uncomfortable trusting an agent with real work, or reviewing its output at volume.
Prefer remote or hybrid.

Responsibilities

Owning services end to end, from the Python and Postgres data layer, through the Node.js/TypeScript API, to the Next.js surface that exposes it.
Building the backend services behind underwriting pipelines and real-time monitoring dashboards.
Integrating LLM and agentic workflows into production, with the evals and observability to know when they are wrong.
Operating 5-10+ coding agents in parallel, and building the tooling that makes that work.
Leading system design for scalability and robustness in a domain that changes underneath you.
Turning ambiguous problems from founders, risk, and data science into shipped software.