Documentary AI-Assisted Production Workflow

You are a documentary post supervisor. Design an AI-assisted workflow for long-form documentary built from huge, multi-year, multi-format interview and archival material. Inputs: - Footage scope: [INSERT - hundreds of hours, interviews, archives, formats] - Story approach: [INSERT - investigative, biographical, observational] - Languages and translation needs: [INSERT] - Archive and rights constraints: [INSERT] - Delivery spec: [INSERT - feature, episodic, broadcaster technical standards] Deliver in modules: 1. Objective: find and shape stories from massive real-world material while preserving journalistic accuracy. 2. Process: ingest/restore and automated logging -> interview transcripts with thematic coding -> semantic story research across footage -> fact/claim verification flags -> stringouts and paper cuts -> archival sourcing and rights tracking. 3. Toolchain: [INSERT - VLM logging, transcription, semantic search, editing tools], fact-check database, rights ledger. 4. QA: transcript accuracy, context-preservation checks on AI-selected quotes, sensitive-content handling, legal and factual review gates. 5. Delivery: selects bins, stringouts, paper-edit documents, rights-clearance reports, and conform handoff. Output: research workflow, coding taxonomy, logging time savings, and editorial ethics guidelines for AI assistance. ================