AI Harness Engineer – AI Chemistry

You've built an agent runtime from the ground up, not just prompts around an API. Own the engine of an AI platform for chemistry.

A Series B company in the San Francisco Bay Area does AI, lab validation and chemical manufacturing under one roof, for more than 100 enterprise customers. Its conversational platform plans how chemicals are made, designs processes and learns from what happens in the lab. You'll build and own the agent harness at its core: the runtime, execution graph, tool contracts, state, evals and reliability in production. It connects scientific engines, reaction databases, simulation software and lab systems, plus the rollout and trajectory plumbing for RL. No chemistry background needed, just the drive to learn it from the experts around you.

What you'll own

• The agent runtime and execution graph
• Tool contracts and integrations with scientific software and lab systems
• State, concurrency and recovery for long-running, multi-step jobs
• Evals and tracing that catch failures on real user tasks
• Rollout and trajectory pipelines for RL and post-training

Must-haves

• You personally created the layer agents run on (an agent harness, orchestration runtime or graph-based framework); prompt engineering, API wrappers or simple RAG alone won't fit
• Solid production Python and real systems experience: managing state, concurrent work, failures, safe retries and jobs that run for hours
• 2+ years of experience, with base salary expectations within the $200k–$300k range
• Working from the San Francisco office five days a week, 9 am to 9 pm (no remote or hybrid)

Bonus points

• An agent runtime you owned at an AI lab or an agentic enterprise AI company
• RL environments, rollout pipelines or scientific tool integration
• Open-source contributions to agent or workflow frameworks, or publications

What's in it for you

• $200k–$300k base + meaningful equity (early-hire packages lean towards equity)
• A founding seat next to AI researchers, chemists and product engineers
• Visa sponsorship available

Good to know

• Full-time, on-site in San Francisco, 9-9-5
• Process: profile review → technical interview with the AI lead → founder conversation → technical leadership interview → references → offer

When you apply, please include

• Your CV (PDF), your LinkedIn URL, and GitHub if you have one
• The agent runtime or harness you built: what it does, the trade-offs you made and what you learned
• Your current location, and confirmation that five days a week in San Francisco works for you
• Your visa or citizenship status
• Your base salary expectation, your notice period, and any other interview processes you're in

Does this sound like you? Apply now. If you're a match, we'll be in touch within 3 working days.

RnB Consultancy | Netherlands | 3604 AB Maarssen | KvK 71489703