kedarg63/hedgefund ? reverse-engineered prompt

Build me an internal research data pipeline for an hedge fund that pulls primary source data only, from India and the US, and stores everything in a raw archive first so nothing gets lost if a source changes later.

I want it to collect market and filings data, like SEC filings and company facts, Indian exchange and regulator data, commodities and RBI publications, then parse it into a warehouse with a clear date stamp for when the information was actually known. Please also add a simple daily runner that can do a dry run, run one job, or run everything due for the day, plus a Streamlit dashboard that shows what was collected and, most importantly, whether the data looks trustworthy.

If it helps, look up current docs online for any tricky APIs or libraries. Keep it practical, with local config from a .env file, basic alerting for failed jobs, and a small API plus web shell for browsing the warehouse internally.