jeffemmett
/
jeffsi-meet
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
21 Commits 2 Branches 0 Tags 134 MiB
Commit Graph

4 Commits

Author SHA1 Message Date
Jeff Emmett f40b8bd97d fix(mi-api): handle RunPod vLLM output format and jsonb serialization 
Fixes three issues preventing summary generation via RunPod GPU:
- Extract text from vLLM's nested choices/tokens output format
- Strip LLM preamble text before JSON parsing
- Serialize lists to JSON strings for asyncpg jsonb columns

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-23 10:14:41 -07:00
Jeff Emmett 76d44975cb feat(mi-api): add configurable timeout + AI orchestrator GPU fallback 
Increase Ollama timeout from hardcoded 120s to configurable 600s default.
Add optional AI Orchestrator integration for RunPod GPU acceleration with
automatic fallback to direct Ollama when orchestrator is unavailable.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-22 20:08:28 -07:00
Jeff Emmett dd12348da8 fix(meeting-intelligence): fix transcription pipeline bugs 
- Fix JSONB serialization in API database.py (json.dumps + ::jsonb casts)
- Fix integer vs UUID job ID handling in transcriber
- Fix UUID-to-string conversion for meeting_id in processor
- Add whisper.cpp shared libraries to Dockerfile (libwhisper, libggml)
- Fix Jibri finalize script log directory path
- Add graceful error handling for speaker diarization
- Support video_path parameter for automatic audio extraction

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
 2026-02-06 12:25:59 +00:00
Jeff Emmett 4cb219db0f feat(meeting-intelligence): add backend infrastructure for transcription and AI summaries 
Add complete Meeting Intelligence System infrastructure:

Backend Services:
- PostgreSQL schema with pgvector for semantic search
- Transcription service using whisper.cpp and resemblyzer for diarization
- Meeting Intelligence API with FastAPI
- Jibri configuration for recording

API Endpoints:
- /meetings - List, get, delete meetings
- /meetings/{id}/transcript - Get transcripts with speaker attribution
- /meetings/{id}/summary - Generate AI summaries via Ollama
- /search - Full-text and semantic search
- /meetings/{id}/export - Export as PDF, Markdown, JSON
- /webhooks/recording-complete - Jibri callback

Features:
- Zero-cost local transcription (whisper.cpp CPU)
- Speaker diarization (who said what)
- AI-powered summaries with key points, action items, decisions
- Vector embeddings for semantic search
- Multi-format export

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
 2026-02-05 19:04:19 +00:00