Databricks has announced the public preview of the Mosaic AI Agent Framework and Agent Evaluation tools, which aim to help developers build and deploy high-quality generative AI applications. These tools address challenges in ensuring AI output is accurate, safe, and governed. They provide features such as human feedback collection, application quality metrics, and an end-to-end development workflow integrated with MLflow. The integrated tools help in managing the lifecycle of AI applications, from development to deployment, and include a Generative AI Cookbook for guidance.
Table of contents
Challenges with building high-quality Generative AI applicationsIntroducing Mosaic AI Agent Framework and Agent EvaluationBuilding a high-quality RAG agentPricingNext StepsSort: