iUser Story:
As a PSAP engineer working on LLM inference performance I want to standardize on:

• Set of models we are focused on evaluating
• How we present the results (shared visualization tooling)
• What metrics we look at and how we measure them. For example, 95th vs 99th percentile ITL. 

So that we can compare the different performance results we have gathered across the team

Acceptance criteria: