1. We ingest market and fundamentals data, then normalize symbols and freshness windows.

2. We compute score components and grade outcomes from predefined rules and threshold bands.

3. We generate cohorts (sector/watchlist/portfolio) for comparative analytics and decision context.

4. We surface alerts and summaries only after policy checks and confidence filters.

This launch version is optimized for clarity and reliability over UI polish.