SmartMeeting

AI Meeting Recording and Task Management System

Quick Start

Docker (Recommended)

docker compose up --build

• Frontend: http://localhost:5174
• Backend: http://localhost:${BACKEND_BIND_PORT:-8000}
• MySQL: localhost:3307

Default account: configured in the seed data

Local Development

One command (Recommended):

npm run dev

This command will:

• install/update backend + frontend dependencies automatically
• start backend on http://127.0.0.1:8000
• start frontend on http://127.0.0.1:5173

Backend:

python -m pip install -r backend/requirements.txt
python -m uvicorn backend.main:app --reload

Frontend:

npm --prefix frontend install --cache "D:\SmartMeeting\.npm-cache"
# 默认代理后端地址为 http://127.0.0.1:8000
# 如需覆盖：set SMARTMEETING_DEV_BACKEND_URL=http://127.0.0.1:8000
npm --prefix frontend run dev

Project Engineering Entry (Recommended)

npm run bootstrap

Unified commands:

npm run dev
npm run dev:backend
npm run dev:frontend
npm run ci

Current Progress

Core Features

• User auth (login/register, JWT, role-based access)
• Meeting CRUD (create, list, detail, filter, search)
• Online recording from microphone (start/pause/resume/stop then transcribe)
• Audio upload + Whisper transcription (mock ASR fallback with realistic multi-segment output)
• AI meeting summary + task extraction (real LLM or rule-based fallback)
• Task management (status transition, priority, due dates, reminder flags)
• Participant management
• Transcript viewing

Frontend Pages

• Dashboard with stats overview and recent meetings
• Login page with JWT auth
• Meeting list with filter/search/pagination
• Meeting detail workbench (summary, transcripts, tasks, create task)
• Task center with status toggle, priority tags, meeting links
• User management with create/delete

Engineering

• Docker Compose one-click startup (MySQL + backend + frontend)
• GitHub Actions CI (backend tests + frontend build)
• Backend test suite (33/33 passing)
• Frontend typecheck + build passing
• Project documentation (backend, frontend, docs READMEs)

Next Steps

• Production environment configuration

Configuration

LLM (AI Summary & Task Extraction)

Without configuration, the system uses rule-based fallback. To enable real AI:

# .env
LLM_PROVIDER=openai
LLM_API_KEY=sk-your-key
LLM_MODEL=gpt-4o-mini
LLM_BASE_URL=https://api.openai.com/v1  # optional
LLM_FALLBACK_PROVIDER=ollama
OLLAMA_BASE_URL=http://host.docker.internal:11434
OLLAMA_MODEL=llama3.1
OLLAMA_TIMEOUT=60
OLLAMA_TEMPERATURE=0.3
OLLAMA_MAX_TOKENS=2000
JWT_SECRET_KEY=change-me-in-production
JWT_ALGORITHM=HS256
JWT_ACCESS_TOKEN_EXPIRE_MINUTES=1440

Whisper (Speech Recognition)

Without Whisper, the system uses mock ASR with realistic multi-segment transcripts. To enable real transcription:

# Install
pip install openai-whisper
# On Windows, also install ffmpeg: https://ffmpeg.org/download.html

# .env
WHISPER_MODEL=base
WHISPER_DEVICE=cpu  # or "cuda" for GPU
WHISPER_LANGUAGE=zh
WHISPER_HOT_WORDS=SmartMeeting,Whisper,WhisperX,faster-whisper,FastAPI,Vue3,TypeScript,Pinia,Vite,MySQL,Docker,JWT,FFmpeg,PyTorch,CUDA,会议纪要,行动项,待办事项,项目推进,接口联调,需求评审,版本发布,任务拆解,进度同步,语音识别,转写结果,摘要生成,热词管理,参与者,截止时间,优先级

If you set WHISPER_DEVICE=cuda, make sure the host has NVIDIA Container Toolkit and a CUDA-capable PyTorch runtime available inside the container.

Documents

• Backend API: docs/backend-api.md
• Frontend runbook: docs/frontend-runbook.md
• Database design: docs/database-design.md
• Engineering framework: docs/engineering-framework.md
• Project planning: docs/SmartMeeting智能会议系统项目立项规划书