AI Engineer

Vibehackers; Chicago, Northern (IL, KY); Hybrid; USD 180,000 - 220,000; Full time

Benefits offered by this job

Equity options

Health, dental, and vision

Parental leave

Flexible vacation

Wellness allowance

Technology allowance

L&D programs

Job summary

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

Responsibilities include end-to-end workflow design, durable-execution maintenance with versioning, and measurable model behavior with tracing and diagnostics within a hybrid startup environment.

Qualifications

• Hands-on experience with LLM agent systems and tool-calling patterns.
• Production service fundamentals: API design, data contracts, authorization boundaries, and observability.
• Fluency in a strictly-typed codebase and strong production ownership.
• 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, and uses evaluations to validate model capabilities.

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.

Skills

LLM agent systems

Tool-calling patterns

Production ownership

Observability

Strong written communication

Startup experience

Safety as code

Tools

Anthropic SDK

MCP servers

Protobuf

Avro

Schema registry

Job description

Uses coding agents and frontier LLMs; builds agentic workflows and AI tooling to make AI coworkers trustworthy in production.

About the Role

Build and operate production AI systems that perform judgment-heavy insurance work, shipping agentic workflows, durable-execution infrastructure, and measurable model behavior. Own design, deployment, and observability for LLM-driven systems in a hybrid, collaborative startup environment.

Job Description

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.
• 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).
• 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
• 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

Anthropic SDK MCP Avro Protobuf Event bus OLTP databases OLAP databases Object storage LLM agent systems LLMs

Skills

System Design Observability Testing API Design Data Contracts Authorization and Security Production Ownership Design Documentation Reliability Engineering Prompt Engineering Evaluation & Metrics Event-driven Architecture Collaboration & Communication Startup Experience

Experience Level

Senior

USD 180,000 - 220,000/year

• Competitive compensation and equity / stock options
• Dental insurance
• Parental leave
• Wellness allowance
• Technology allowance
• Company-sponsored personal and professional development
• L&D partnerships and monthly Lunch & Learns
• Wellbeing perks (Peloton, Fetch, OneMedical, Headspace care+)
• Caregiver support (dependent care FSA seed, Care.com membership)