Senior Software Engine

SENIOR SOFTWARE ENGINEER, AI AGENTS

This role is five days a week onsite in San Francisco. If that works for you, and you have strong software engineering fundamentals and real curiosity about agentic AI, it is worth reading on.

The company is a funded AI software business in San Francisco building automation tools for enterprise software customers. The product is live, in production, and used daily by thousands of paying users.

The role

This is a senior individual contributor role on the agent itself, the core system the product is built around. You will work directly with the technical founder and a small team of engineers, owning real pieces of the agent end to end. Agentic AI experience is useful but not required. What the team needs is strong software engineering fundamentals and genuine interest in going deep on this problem.

Why it's worth a look

The agent is the product. Every other system at this company exists to support it, which means the work here is central, not peripheral. The product is in production with real customers and a real feedback loop, so what you build lands quickly and you can measure whether it works. You will report directly to the technical founder and have genuine input into how the agent is designed, measured and improved. The team is intentionally small, which means ownership is real and decisions move fast. For an engineer who wants to go deep on agentic systems at a company where that is the whole problem, not a feature, this is a rare seat.

What you'll be doing

- Build and own core parts of the agent system, including how it handles state across long workflows, recovers from errors, manages context, and routes between models

- Measure what matters, task completion rates, cost per workflow, token efficiency, and build the evaluation infrastructure that tells the team whether the agent is actually improving

- Write production code that runs in real enterprise software environments, where robustness, concurrency and failure handling matter as much as correctness

- Work with a user researcher and domain specialists to understand how customers actually work, then translate that understanding into benchmarks and test cases the agent is measured against

Must have:

- Three or more years of commercial software engineering experience, shipping and maintaining production systems

- Strong computer science fundamentals, data structures, algorithms, systems design

- Solid Python skills and meaningful depth in at least one other language

- Ability to work onsite five days a week in San Francisco

Nice to have:

- Experience with agentic AI or LLM-based systems

- Background at a high-growth startup

- Familiarity with how software behaves in locked-down enterprise environments

- Experience with desktop automation or programmatic control of third-party applications

Package

- $180,000 to $330,000 base salary, depending on experience and interview performance

- Onsite, San Francisco, five days a week

- Competitive equity included, four-year vesting schedule

- Relocation support available for candidates moving to the Bay Area

- Visa transfers and new sponsorship supported

Get in touch with Daryl Crothers at Oxbow Talent to find out more about the role and the team.