Founding Harness Engineer – AI Agents

This is a fully remote position, open to applicants in United States.

📋 Description

• Develop the agent runtime, which includes context management, tool selection, approval processes, and decisions regarding when to act, inquire, or remain silent.

• Continuously write and refine system prompts.

• Take ownership of the evaluation and regression framework, testing prompt modifications similarly to code.

• Ensure long-running tasks are resilient against interruptions, retries, and unreliable APIs.

• Implement safety measures, approval checkpoints for high-risk actions, and maintain strict separation of customer data.

• Collaborate directly with the founder and early users, translating feedback into actionable features.

⛳️ Requirements

• Successfully deployed tool-utilizing LLM agents to actual users in a production environment; demos and prompt wrappers are not considered sufficient.

• Strong instincts for agent initiative, validated by testing outcomes.

• Practical experience in constructing evaluation and regression suites for LLM performance.

• Knowledge of safe tool execution, prompt injection vulnerabilities, and risks associated with multi-tenant data.

• Proficient backend development skills in Python or TypeScript, including handling asynchronous tasks, retries, and state management.

• Capability to transform an ambiguous problem from prototype stage to a production-ready solution while articulating the associated trade-offs.

• Exceptional written and verbal communication skills.

• Nice to have: Experience with MCP or integrations with Slack, Gmail, Notion, or Linear.

• Nice to have: Familiarity with evaluation and tracing tools such as Braintrust, Langfuse, or LangSmith.

• Nice to have: Background as an early engineer at a startup.

🏝️ Benefits

• Competitive salary plus equity.

• A founding position with ownership of the agent framework and a significant influence on product functionality.

• Direct access to the founder and early users, without intermediary layers.

• Fully remote work opportunity anywhere in the US, preferably with hours that align with Pacific time.

• Access to the necessary tools and models to execute the role effectively.

• A small, skilled team that empowers you to manage your own tasks.

• Potential pathway to full-time employment once mutual agreement is established.