Mosaic AI Agent Framework: Creating Quality AI Agents

In this chapter, we will discuss the secret weapon for updating a large language model (LLM) with custom unstructured data, like PDF or PowerPoint. While most applications allow you to build a bot or GPT very easily, enterprises are looking for ways to evaluate the quality of the chatbot. This is where the AI Agent Framework comes in. We will not only discuss how to deploy a chatbot from end to end, but how to evaluate it with an LLM as a Judge or human feedback. These metrics will ensure data scientists who are already familiar with MLflow will be able to transition to LLM evaluation easily.