About the Role
Build the production infrastructure that helps AI agents execute tasks reliably, retain useful context, and involve people when needed. You will work across agent systems, evaluation, and operational workflows, shipping improvements informed by real-world use.
What You'll Do
Build and maintain infrastructure that enables agents to complete tasks reliably across many iterations.
Design memory systems that retain client context from messy operational data over time.
Develop evaluation harnesses that teams can use to make production deployment decisions.
Measure, test, debug, and improve agent systems, including their ability to learn from failures.
Ship agent systems and iterate based on production feedback.
Contribute across infrastructure, orchestration, customer collaboration, and team growth as needed.
What We're Looking For
At least 3 years of experience building and shipping LLM agents that ran unattended in production for real users.
Strong Python skills and hands-on experience building production systems.
Experience implementing memory and context retrieval for real operational data, beyond demo RAG projects.
Experience with agent orchestration, infrastructure, workflow management, and production reliability patterns such as retries and failure recovery.
Built evaluation harnesses for LLM systems that informed production deployment decisions.
Experience building agent tools, frameworks, or libraries, or leading a technical project from conception to production.
Open-source contributions to agent tooling or LLM infrastructure are a plus.
Compensation & Benefits
Base salary is $150,000 to $250,000 USD annually, with meaningful early equity. Visa sponsorship and relocation support are available.
Location
On-site in San Francisco, United States.