One of the key problems we are solving at Composio is self improve tools, and for this we are hiring a Research Engineer to work on evals. What are we trying to evaluate here exactly - was the agent able to complete its desired action? If not, then what caused it? By action we mostly mean tool action here, but this would be the high level problem statement.

This role will be largely a self exploratory role with broad autonomy so we except someone who works well under those conditions and has agency.

This role is can be completely remote but you are welcome to work out of our SF or BLR offices.

If you are exceptional and we want to work together we expect pay to not be an issue

How to apply
Email hiring@composio.dev - demonstrate your excellence