v4.2.0 - AI Provider Abstraction & BYO-AI
P1. No commitment to any single ASR or LLM provider. A normalized contract (async per-track ASR with word timestamps, json_schema LLM output, declared context windows and capability flags) with swappable adapters, per-campaign BYO-AI, and SELF-HOSTED AS A FIRST-CLASS TARGET on both CPU and GPU — not a fallback. Chunk sizing is driven by the provider's declared window, so one code path is correct on a 9B local model and a 1M-context hosted one. Supersedes and absorbs #128. See section 6.
No due date
100% Completed
