About the Role
Join the engineering team at an agentic integration platform for real-world operations and own the agent layer end to end. You will build production systems that help agents understand undocumented environments, recover from failures, and improve through accumulated knowledge.
What You'll Do
Build context and self-learning systems that improve agent performance as they encounter more integrations.
Develop self-healing workflows that detect production failures, diagnose causes, and recover autonomously.
Own the agent harness and meta-harness, enabling teams to build, evaluate, and evolve agents quickly.
Set engineering patterns for building, measuring, and improving agents across the platform.
Work closely with company leadership, experiment rapidly, and help grow the team responsible for agentic systems.
What We're Looking For
Two to nine years of professional software engineering experience, with recent experience building agentic software.
A completed or in-progress engineering, computer science, or quantitative degree from a top-250 technical university; a top-50 university is strongly preferred.
Experience as a founder of a startup that raised outside funding or hired a full-time employee, or as an engineer at a pre-seed, seed, or Series A startup backed by a top-tier venture firm.
Hands-on experience with context engineering, reusable agent harnesses, agent tracing, evaluations, and production failure detection, diagnosis, or recovery.
Experience shipping working production systems, rather than research alone. Familiarity with TypeScript, Effect-TS, PostgreSQL, LLMs, and agent evaluation is relevant.
Compensation & Benefits
Base salary is $150,000 to $300,000 USD annually. Visa sponsorship is available.
Location
On-site in San Francisco, California, five days per week. Candidates based elsewhere in the United States should have a concrete, dated relocation plan to San Francisco.