Evaluate Agent Responses for GenAI Applications > Introduction to Evaluate Agent Responses for GenAI Applications recipe
Introduction to Evaluate Agent Responses for GenAI Applications recipe
The Evaluate Agent Responses for GenAI Applications recipe is based on REST and SOAP APIs. Use the recipe to evaluate the quality of responses from the Large Language Model (LLM) based on the specified evaluation criteria, ensuring a structured and objective assessment.
You can complete the required fields in the Evaluate Agent guide and upload a file containing grounded data, with a maximum size of 5 MB, in any of the following formats:
.txt, .xml, .json, .html, .csv, .md
The process extracts data from the uploaded file to create payloads for the first LLM based on the user prompt, response instruction, and the grounded data.
The process then creates a payload for the second LLM based on the user prompt, evaluation instructions, grounded data, and the response from the first LLM. The process then parses the JSON response from the second LLM and creates output in a specific format.