About the role
SynchrotronAI® builds governed AI agents that do real work: Tariff Agent, Capital Agent, ScamFreeLife, and AI Off Now™.
You will work with our Head of Research on the agents behind these products: how they reason over documents, use tools, stay within governed boundaries, and get measurably better with every release.
What you will do
- Design and build AI agents that read documents, use tools, and complete multi-step work.
- Build evaluation suites that measure accuracy, safety, and cost, and use them to guide every change.
- Test new models as they are released, learn their strengths and limits, and decide where they fit.
- Work across model providers and techniques: prompting, retrieval, structured outputs, and model customization.
- Make agents fast, reliable, and affordable in production.
- Build the guardrails and human-review steps that keep AI under human control.
- Turn research ideas into product features with the founders.
What you bring
- A degree in computer science or a closely related field.
- Three or more years of software engineering experience.
- Strong Python and TypeScript, and working knowledge of Rust.
- Hands-on experience building production applications with large language models.
- Experience designing evaluations and measuring model quality.
- Sound engineering judgment about reliability, cost, and security.
Nice to have
- Research experience in machine learning, natural language processing, or AI safety.
- Experience with agent frameworks, tool use, retrieval systems, and the Model Context Protocol (MCP).
- Experience with model inference, serving, or GPU workloads.
- Experience with C++, for example in model inference or other performance-critical code.
- Publications, open-source work, or competition results.
- Experience with document-heavy domains such as trade, lending, or government.
Don't meet every qualification? If the work excites you, we'd still like to hear from you.
Compensation and benefits
Competitive pay.
- Remote, anywhere in the U.S.
- Equity
- 401(k)