This project builds an agent-based time series forecasting system that automates the entire modeling pipeline from raw data to final insights. Users upload a dataset, and the system uses multiple AI agents to understand the data, preprocess it, select appropriate models, train them, and evaluate performance. It then generates predictions, compares models using metrics like MAE and RMSE, and presents results through an interactive dashboard. The goal is to make forecasting faster, more accurate, and accessible by removing manual steps and enabling end-to-end automation.