Staff AI Engineer - Agent Architecture & Behavior

Artisan
San Francisco
Workplace: OnsiteFull timeUSD 250,000 - 325,000 annuallyFunction: Architecture & Urban PlanningExperience: 3+ yearsSkills: ["Technical ownership","Technical leadership","Debugging","Experimentation","Collaboration"]

Design and build the core architecture for agentic AI systems, owning execution loops, planning, tool interfaces, and verification. You’ll develop reliable multi-agent workflows with resilient stateful execution, context/memory improvements, and reusable methods. Partner with engineers to deliver production services, integrations, secure execution, and observability. Lead technical standards and transform real-world failures into measurable evaluation and experimentation improvements.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Artisan
Artisan
6 months ago

Staff AI Engineer - Agent Architecture & Behavior

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 40 minutes agoStatus: Live

Job Summary

Design and build the core architecture for agentic AI systems, owning execution loops, planning, tool interfaces, and verification. You’ll develop reliable multi-agent workflows with resilient stateful execution, context/memory improvements, and reusable methods. Partner with engineers to deliver production services, integrations, secure execution, and observability. Lead technical standards and transform real-world failures into measurable evaluation and experimentation improvements.
Location: San Francisco
Workplace: Onsite
Employment Type: Full time
Job Function: Architecture & Urban Planning
Seniority: Mid level

Key Responsibilities

  • •Design and implement agent execution loops, planning strategies, tool interfaces, and verification, turning ambiguous requirements into working code and clear system boundaries.
  • •Build multi-agent systems for delegation, coordination, context sharing, and result synthesis, handling concurrency, conflicting updates, cancellation, and stale results.
  • •Create resilient execution for complex stateful workflows, including checkpoints, recovery strategies, and appropriate human intervention.
  • •Improve context, memory, retrieval, persistent state, and skill representation to improve performance without regressions.
  • •Build evaluation and experimentation to analyze failures and compare approaches on quality, reliability, latency, and cost.

Pay and Benefits

Salary: USD 250,000 - 325,000 annually
Equity and Bonus:Equity
Perks:Equity

Key Requirements

  • •Have personally built and shipped a substantial agentic system, with production use or rigorous, reproducible open-source work.
  • •Have deep practical experience with LLM tool use, planning, context engineering, and evaluations, including multi-agent coordination or parallel agent/tool execution.
  • •Have hands-on experience with browser or computer automation in an agentic system, including observing state, verifying effects, and recovering from failures.
  • •Be an excellent software engineer in Python, TypeScript, or a comparable language, comfortable with asynchronous services, persistence, concurrency, retries, and cancellation.
  • •Know which guarantees must be enforced in code vs. left to the model, and can reason about permissions, untrusted inputs, and human approvals.
Experience:3+ yearsAgentic AIMulti-agent systemsLLM tool useOpen source
Skills:Technical ownershipTechnical leadershipDebuggingExperimentationCollaboration
Tech Stack:PythonTypeScriptLLMs

Eligibility

Work Authorization:Sponsorship available.

Company Brief

Artisan
Operates a curated talent marketplace that connects companies with vetted freelance and contract software engineers, designers, and product builders to staff projects and augment engineering teams.
Industry: Recruitment Agencies
Website