Harness Engineer

Are you passionate about pushing the boundaries of AI and agentic engineering? This role offers a unique chance to shape the future of autonomous systems. Imagine working on the cutting edge of AI, where your innovations directly impact the efficiency and intelligence of agents used by thousands of workspaces. If you have a proven track record in building or significantly modifying AI agent harnesses, this could be the perfect fit for you.

Key duties and responsibilities:

- Build the agent runtime: Develop the loop where the agent plans, acts, and verifies its own work.

- Engineer the context: Shape how system prompts, skills, memory, and tools interact on complex problems.

- Enhance tool fluency: Integrate various tools and ensure the agent can recall and use the right functions.

- Expand agent capabilities: Develop reliable memory, self-written skills, and autonomous work execution.

- Improve performance: Conduct evaluations, catch regressions, and turn failures into fixes.

- Ship product features: Deliver features to users via platforms like Slack and Teams within hours.

Skills and experience:

- Proven experience in building or modifying AI agent harnesses.

- Deep understanding of how system prompts, skills, and tools interact in complex scenarios.

- Experience in creating custom skills, commands, CLIs, or MCP servers for agentic coding.

- Proficiency in agentic coding and stress-testing models and harnesses.

- Strong systems thinking with the ability to make technical trade-offs and design for scale.

- Fast-paced work ethic with a focus on immediate shipping.

- Genuine interest in AI, backed by reading relevant papers and forming opinions on evaluations.

Why this role is different:

- Direct collaboration with founders, ensuring swift decision-making.

- The agent is the core product, with every reliability improvement directly impacting user retention.

- Frontier work in building a harness at the edge of current AI capabilities.

Even better if:

- Previous AI engineering experience with agents, harnesses, or evaluation infrastructure.

- Experience as a founder or early-stage builder.

- Contributions to open-source AI tooling.

Tech:

- Python expertise required.

- No reliance on existing agent frameworks or orchestration libraries; build the harness from scratch.

How the team works:

- Small, high-trust team with low process overhead.

- Immediate impact: ship your first week, engage with users from day one.

- No alignment meetings or stakeholder syncs; focus on building, testing, and iterating.

Why this role stands out:

- Singular focus on building a harness for AGI-level models.

- Rapidly growing company with a product that works and a market that demands it.

- Unique ownership opportunities at this stage, giving you a significant surface of the company to shape.

Compensation:

- Top-of-the-market salary and ownership opportunities.

- Remote-first role with hubs in Munich, New York, and Warsaw, and occasional in-person team gatherings.

Ready to take your AI engineering skills to the next level? Apply now and be part of a team that's redefining what's possible in autonomous systems.