Today we are launching some new functionality that makes it easy to connect your Langfuse data and pull it in to Neurometric for analysis. Simply enter your Langfuse information and we can take your AI prompts and analyze them on thousands of system combinations to tell you what models and test time compute algorithms work best to minimize cost or latency at a given level of accuracy.
The screenshot below shows a sample analysis from a Langfuse implementation and evaluation run. Using the langfuse data it shows which models you should move to if you want to optimize your AI system.
If you want help using Langfuse in Neurometric, please reach out.