rishu765/mna_deal_intelligence ? reverse-engineered prompt

Build me a Python portfolio for AI assisted M&A deal intelligence that walks through the whole workflow from company research to target screening, valuation, precedent transactions, diligence, and final deal synthesis.

I want six connected parts, each one able to run on its own, but also work together through a single orchestration layer. The app should be able to take source documents, search and summarize evidence with citations, rank acquisition targets from a thesis, estimate value using public comps and precedents, review diligence risks from a VDR style document set, and then combine everything into a grounded analyst friendly output.

Please make it work offline with deterministic demos and clear human review steps, so it does not behave like an autonomous decision maker. Keep the finance math explicit and separate from the model reasoning, and make sure results show their supporting evidence and any warnings. If you need to check current docs or best practices while wiring it up, go ahead and look them up online.