AI Engineer

Role

Vouch is hiring an AI Engineer to build AI software that performs judgment-heavy insurance work. You’ll design and ship agentic workflows end-to-end, extend a durable-execution substrate, and maintain measurable model behavior and observability for systems that run in production and interact with real users.

Key Responsibilities

• Ship agentic workflows end-to-end: design tool contracts, capability boundaries, and approval gates, then carry changes through review, deploy, and production ownership.
• Extend and maintain a durable-execution backbone with pinned worker versioning and replay compatibility so crashed workers can resume mid-workflow without losing state.
• Make model behavior measurable: maintain deterministic corpus tests as merge gates, build eval suites and diagnostics, and maintain LLM tracing and token/cost observability.
• Treat prompts and tool definitions as engineered artifacts: version, snapshot-test, cache-stabilize, and review them like code.
• Run the event and data substrate: event bus with schema-governed domain events, object storage, OLTP/OLAP databases, and two-way plumbing between agents and systems of record.
• Use and build with frontier coding agents as daily instruments, and build the systems that make AI coworkers trustworthy.
• Participate in a strong review culture that focuses on finding silent failure modes before production.

Requirements

• Hands-on experience with LLM agent systems and tool-calling patterns; familiarity with MCP, the Anthropic SDK, or equivalents running in front of real users.
• Production service fundamentals: API design, data contracts, authorization boundaries, and observability.
• Fluency in a strictly-typed codebase and strong production ownership (debugging OOMs, correctness reworks, etc.).
• Strong written communication: design docs, review comments, and public reasoning channels.
• Experience in early-stage startups or high-growth environments and a bias toward small, traceable increments of delivery.
• A mindset that treats safety properties as code, not just prompts, and uses evaluations to validate model capabilities.

Nice-to-haves

• Experience with durable-execution engines in production.
• Experience with event-driven systems with schema governance (pub/sub, schema registry, Avro/Protobuf).
• Familiarity with eval frameworks, LLM observability, MCP servers, and data lake/warehouse adjacent engineering.
• Prior work in regulated domains (insurance, fintech, healthcare).
• Frontend experience.

Work Environment

• Vouch is a U.S.-distributed team with offices in San Francisco, Chicago, and New York City. This role has hybrid work flexibility and requires being in the office at least three days per week (Tuesday–Thursday).

Benefits

• Competitive compensation and equity / stock options
• Health, dental, and vision insurance
• Parental leave
• Flexible vacation time
• Wellness allowance and technology allowance
• Company-sponsored personal and professional development; L&D partnerships and monthly Lunch & Learns
• Wellbeing perks (examples: Peloton, Fetch, OneMedical, Headspace care+)
• Caregiver support (dependent care FSA seed and Care.com membership)
• Regular performance reviews and development discussions

Compensation & Location

• Listed salary range: 180,000 - 220,000 USD per year
• Location example provided: Chicago, IL (Hybrid)

Interview Process (typical)

• 30-minute recruiting phone screen
• 30–45 minute video interview with hiring manager
• Case study / technical screen
• 30–45 minute 1:1 video interviews with 3–4 team members
• Executive chat