Curated by EarlyCove

Tools for evaluating and observing AI applications

Find prompt, trace and evaluation tools for understanding failures and improving model-powered products over time.

How to choose

Instrument one application path and turn an observed failure into a repeatable evaluation. Compare how each tool connects versions, traces and datasets. Check what data is captured, how it is retained and whether a gateway is part of your architecture.