I need an interactive dashboard built in Streamlit that lets end-users explore time-series data coming from three different sources—raw CSV uploads, existing relational databases, and live API endpoints. The app should read, clean, and merge these feeds on the fly, then offer clear visual insights through line charts, area charts, and any other plots that make trends, seasonality, and anomalies obvious. Under the hood I expect well-structured, reusable Python code that leans on pandas for manipulation, SQLAlchemy (or similar) for database access, and a lightweight requests layer for the APIs. Caching, session-state handling, and responsive layout controls are important so the interface feels fast even as data volumes grow. Deliverables • Streamlit app folder with modular, commented Python scripts • A config file (YAML or .env) that lets me switch between data sources without touching code • README with setup instructions and screenshots of the finished dashboard • Short video or GIF showing the main interactions, proving everything works end-to-end I’ll test by pointing the app at sample CSVs, a Postgres sandbox, and one public API. If each source loads and visualizations render correctly without errors, the job is complete.