We are looking for a Founding Engineer to join a dynamic team focused on building infrastructure to evaluate and improve AI agents in real-world systems.
What you'll do:
- Turn real failures into tasks that can be scored.
- Build datasets, gold labels, graders, and human review loops.
- Develop methods to prevent reward hacking.
- Analyze agent traces to identify limitations of large language models (LLMs) in production environments.
Who you are:
- Someone who has shipped AI agents or built evaluations or graders for LLMs.
- Enjoys digging through agent traces to find critical insights.
Location: Remote (India preferred)
This is an opportunity to work on challenging AI problems and influence how the effectiveness of AI agents is measured.
