docs: Cluster A (count drift) + Cluster B (deleted-module refs) sweep

fix(batch): voice interim, TTS truncation, RULE2.5 re-ask, faithfulness regression, markdown, compare-reviews, policy-name

fix(#55+#56+#53/#54-engine): TTS full natural readout + voice warm-stream/pre-roll

fix(#40+voice+upload): SSOT scorecard parity + Sarvam 30s-truncation + uploaded-PDF retrieval gate

chore(cleanup): purge stale narrative/tombstones/dead code — codebase reads as the current standard

fix(single_brain): restore broken core flow — model revert + thinkingConfig + rip deterministic scaffolding

recovery: integrate stalled-session work + de-stale/cleanup (pytest 215 green) [build-fix]

data+scoring: verbatim-source all policy_facts, recalibrate scorecard, fix recommendation

fix(ux+voice): KI-274 — landing page enterprise redesign + Sarvam 429 friendly error

refactor: KI-268 — Path B cleanup: -5,219 LOC legacy + nim_fallback + chain collapse + modal wiring

fix(arch+wiring): KI-237..KI-241 — Phase Y bundle (Y1+Y2+Y3+Y4 + profile-gate)

fix(arch+robustness): KI-225..KI-235 — Path B single-LLM rewrite + audit bundle

feat(voice+session+llm): KI-195 + KI-196 + KI-197 + KI-199 bundle

feat(llm+docs): KI-177 + KI-179 + KI-183 + ADR-040 docs cascade

fix(llm): KI-175 — demote nemotron to last resort in all chains

fix(voice+chat+llm+docs): KI-155 + KI-156 + KI-157 + KI-158 + KI-159 + KI-160 bundle

perf(ingest): KI-125 — bump MPS embedding batch 64→128 for bulk re-ingest

fix(latency): KI-099 — bound every per-turn LLM call

feat(llm-chain): KI-088 — cap NIM outbound concurrency at 2 + serialise probe + drop inner retry

docs: ADR-032 LLM chain architecture reference + KI-080..KI-087 doc sync

feat(llm-chain): KI-085 — proactive credit tracking; election gates by quota not just liveness

feat(llm-chain): KI-084 — cut probe token consumption 50× + 429-demote for 1h + per-phase httpx timeouts

docs: KI-080 — sync CLAUDE.md + ENTERPRISE_AUDIT + new ADR-031 + 11 stale KI-025/50-50 surface sweep

feat(llm-chain): KI-080 — sticky primary election; 1-2 LLM calls per turn instead of 5-6

feat(llm-chain): KI-079 — Groq earlier in FAST_BRAIN_CHAIN + escalate to BRAIN_CHAIN on timeout before canonical fallback

feat(llm-chain): KI-078 — tighten per-link timeout + narrow exception catch + stamp fallback reason for telemetry

docs: KI-071 — reconcile docs with code reality (Qwen 80B / Nemotron 30B / Mistral 675B as current primaries)

refactor: KI-050 — complete data/ → 40-data/ rename across all Python refs

refactor: KI-047 — bucket reorg (Option A safe subset)

feat: KI-044 PCM pre-roll via AudioWorklet + 11 folder READMEs

perf: KI-034 retrieval cache + KI-035 reorder FAST_BRAIN_CHAIN

perf: KI-025 — load-balance brain calls across NIM Qwen ↔ Groq Llama

fix: KI-018..022 — routing, profile telemetry, NIM timeouts, judge JSON, chat clear

fix(stt): transcode browser webm/opus → wav before Sarvam (D-024)

Deploy: Stack A (NIM brain + Maverick judge + Sarvam voice). D-019.

Deploy: Stack A (NIM brain + Maverick judge + Sarvam voice). D-019.

Deploy: Stack A (NIM brain + Maverick judge + Sarvam voice). D-019.

Deploy: Stack A (NIM brain + Maverick judge + Sarvam voice). D-019.

Clean Space: delete 263 legacy files (4 retired providers + corpus/vectors/extracted moved to dataset). D-019 Stack A.

Stack A: NIM brain + Maverick judge + Sarvam voice/Indic (D-019). Data moved to insurance-bot-data dataset; Space is now code-only.

Deploy v1 — single-Docker FastAPI + Next.js + RAG + voice + faithfulness

Deploy v1 — single-Docker FastAPI + Next.js + RAG + voice + faithfulness