Summary
Unify builds observability and evaluation infrastructure for AI products. Founded by Daniel Lenton, who has a PhD in ML for 3D Vision and Robotics from Imperial College London, they solve the problem of overly complex and rigid LLM observability tools. Instead of forcing developers into predefined schemas, they offer a simple logging primitive (unify.log) that stores any data type, which can then be visualized using three customizable building blocks: Tables, Views, and Plots. This lets teams monitor production traffic, create datasets, and run evals tailored to their specific product needs.
