Smarter Model Tuning: An AI Agent with LangGraph + Streamlit That Boosts ML Performance

An AI agent is built to provide actionable suggestions for improving machine learning model performance. The agent accepts a model type, either classification or regression, along with its performance metrics to generate tuning advice. This project utilizes LangGraph to structure the agent's workflow as a graph, Google's Gemini model for generating the suggestions, and Streamlit to create the user interface. The architecture involves distinct nodes for handling user input and conditional logic that directs the query to a specialized LLM prompt based on the model type.