Mia Implementation Roadmap

Full legacy body rendered for humans

Mia Implementation Roadmap

Phase 0: Recovery And Storage Decision

• inspect archived Mia code and tests
• identify current canonical content asset table
• map old content_pieces/content_production concepts to current RedKey/B2BEA schema
• decide whether Mia runs as a RedKey role worker, a service called by Jess, or both

Exit criteria: storage and runtime boundary are explicit.

Phase 1: Port Podcast Transcript Processor

• port models.py
• port depth_classifier.py
• port linkedin_composer.py
• port podcast_processor.py
• port tests and update imports
• create fixture transcript test

Exit criteria: transcript in, LinkedIn draft + quote-card brief assets out.

Phase 2: Content Asset Writer

• adapt content_writer.py to current schema
• preserve source tracking and duplicate prevention
• write assets as draft/review_ready
• return asset IDs to caller

Exit criteria: Mia outputs are durable and queryable.

Phase 3: Jess Handoff Integration

• define Jess task payload for Mia
• process transcript from Jess handoff
• return assets to Jess for content kit and approval summary
• block on missing transcript or ambiguous guest identity

Exit criteria: Jess can call Mia for post-recording content assets.

Phase 4: Broader B2BEA Content Modes

• support non-podcast source material
• support articles, carousels, email copy, and resource summaries
• add channel-specific content contracts

Exit criteria: Mia is useful beyond podcasts without weakening podcast flow.