2-4 years of engineering experience with production LLM agents or workflows.
Experience in building agents with tool use, orchestration, and context design.
Strong Python skills and familiarity with PostgreSQL and asynchronous systems.
Experience in writing evaluations and reviewing AI-generated code.
Strong judgment in context management and task delegation.
Familiarity with durable workflow engines like Temporal is a plus.
Authorized to work in the U.S. without visa sponsorship.
Responsibilities
Design and own the agent framework for farm operations.
Build memory systems to capture undocumented farm-specific knowledge.
Develop code automation for agents to handle recurring tasks.
Create evaluations and safety measures for agent reliability.
Integrate machine learning insights into the agent framework.
Observe and refine the product based on farm team feedback.
Benefits
Healthcare coverage.
Access to AI coding tools and models.
Provision of work hardware.
Full Job Description
About the Role
Join the founding engineering team building an AI agent framework for farm operations. You will work closely with AI leadership to design and ship agents that support real business decisions across multiple farms. What You'll Do
Design and own the agent framework, including roles, tools, permissions, approvals, and a runtime for tasks that span multiple days.
Build memory and context systems that capture farm-specific knowledge, including operational knowledge that may not be documented.
Develop code automation that lets agents create and reuse code for recurring work, with human oversight when unexpected situations arise.
Create evaluations and safety measures to improve agent reliability and demonstrate ongoing performance improvements.
Build tools on top of the data layer and integrate machine learning insights into the agent framework.
Observe farm teams using the system and refine the product based on their feedback.
What We're Looking For
Two to four years of engineering experience, including shipping production LLM agents or agentic workflows that people rely on.
Experience building agents with tool use, orchestration, context and memory design, and evaluation.
Strong Python skills and familiarity with PostgreSQL, asynchronous systems, and workflow engines.
Experience writing evaluations as part of agent development, critically reviewing AI-generated code, and independently handling system design.
Strong judgment around retrieval, context management, and when work should be handed to a person or implemented with conventional code.
Experience with Temporal or similar durable workflow engines, agent observability and tracing, or security measures for prompt injection and permissions is helpful.
Authorized to work in the United States without visa sponsorship, and able to work on-site in San Francisco and travel to farms for about one week per quarter.
Compensation & Benefits
Base salary is $170,000 to $200,000 annually, plus 1% equity with four-year vesting and a one-year cliff. Benefits include healthcare, AI coding tools and model access, and work hardware. Visa sponsorship is not available. Location
On-site in San Francisco, five days per week, with about one week of quarterly travel to farms in Washington, Oregon, and California.