Impax Recruitment / Confidential client search

← All opportunities

Agentic AI / ML

Applied AI Engineer

Engineer AI agents that can reason over business data and carry out operational tasks with appropriate safeguards.

Apply for this role ↗
01ROLE OVERVIEW

What you could build.

Impax Recruitment is working with a confidential enterprise software client modernising operational systems for advanced manufacturing. You will build the AI capabilities behind conversational interfaces and multi-step business workflows.

The work

  • Develop agent workflows that maintain state, suggest actions and execute tasks.
  • Connect models to tools, memory and retrieval across structured and unstructured information.
  • Design recovery paths for tool failures, missing context and uncertain outputs.
  • Build evaluations and use the results to improve quality and reliability.
  • Work with frontend and backend engineers to balance responsiveness, model capability and user control.

What you bring

  • Around 4–10 years in backend or full-stack engineering, including bringing AI capabilities into production.
  • Strong engineering fundamentals and experience integrating models with services and data.
  • Care for sensitive information, predictable failure handling and measurable system behaviour.

Technical environment

Python, FastAPI, PostgreSQL, GraphQL, LangChain, LangGraph and AWS model services

Package and working arrangement

  • Full-time, with a hybrid arrangement in California.
  • San Diego-based employees attend three days each week; Los Angeles or Bay Area employees attend for one week each month.
  • Compensation range of $160,000–$250,000 plus equity.
  • Visa sponsorship and visa transfers are not available.

Apply through Impax

Quote IMPAX-104 when contacting Impax Recruitment. We are the recruitment agency representing the client; the employer’s identity is shared during the recruitment process.

YOUR NEXT CHAPTER

Explore the opportunities
behind the headlines.

Discover roles shaped around ambitious teams and real technical challenges.

More opportunities