Abhigna13/Weather-Intelligence-AI ? reverse-engineered prompt

Reverse engineered prompt

Build me a weather intelligence app that can both predict weather and answer weather related questions in a smart chat style assistant.

I want a Streamlit front end for people to explore predictions, see model performance, and view a prediction history, plus a FastAPI backend with a simple health check and a predict endpoint. The app should support machine learning and forecasting for weather data, show basic analytics and explainable insights like feature importance, and include a few time series approaches so the user can compare results. It should also have a research assistant that looks up weather knowledge from a local document set, retrieves the most relevant context, and uses a local LLM to generate answers. If possible, add a simple agent that can decide whether a question should go to the prediction system or the knowledge assistant.

Please make it work end to end, keep it runnable locally, and use current docs online if you need to.

Are you gonna build this?

make sure you review the code using arcumet

Try freeSponsored — opens Arcumet in a new tab