Real-World Examples of LLM Evals Used at Bazaarvoice

AI's transformational potential has permeated our zeitgeist: everyone expects it will change the way we work and the way we live. As AI practitioners, it's our job to activate the change that manifests this destiny. But... what about our lives? In this talk, Lou Kratz--Principal Research Engineer at Bazaarvoice--dives deep into Evals (LLM prompts that evaluate other models)and how their use can speed up your work and transform the amount of time and effort you spend on validation. He does this by sharing real-world examples of Evals used at Bazaarvoice, and shows how Arize products and services can integrate well in your Eval development cycle to speed you up extensively. This talk was originally delivered at Arize:Observe 2024 at Shack 15 in San Francisco on July 11, 2024.