amankumarranjan8969/jal-setu-ground-water-forecasting-and-rag-assistant

Jal Setu is a Streamlit dashboard for seasonal groundwater forecasting, with a question-answering assistant grounded in both run results and project knowledge. The main workflow is upload a seasonal workbook, prepare and model the data, inspect forecasts and diagnostics, and optionally export results. A separate model family adds physics-informed, reservoir, SARIMA, and optional LSTM results. The chat assistant handles small talk directly or answers substantive questions using retrieved documentation and live run context, with optional hosted LLM providers. Some wiring details are inferred only from module descriptions and the partial app excerpt.