maternal-app

andrei/maternal-app

Fork 0

Files

History

Andrei 075c4b88c6

CI/CD Pipeline / Lint and Test (push) Has been cancelled

Details

CI/CD Pipeline / E2E Tests (push) Has been cancelled

Details

CI/CD Pipeline / Build Application (push) Has been cancelled

Details

feat: Add AI streaming responses foundation (partial implementation)

**Streaming Infrastructure**
Created foundation for real-time AI response streaming using Server-Sent Events (SSE):

**1. StreamingService** (163 lines)
- Azure OpenAI streaming API integration
- SSE stream processing with buffer management
- Chunk types: token, metadata, done, error
- Response stream handling with error recovery
- Timeout and connection management

**2. AI Controller Streaming Endpoint**
- POST /api/v1/ai/chat/stream
- SSE headers configuration (Content-Type, Cache-Control, Connection)
- nginx buffering disabled (X-Accel-Buffering)
- Chunk-by-chunk streaming to client
- Error handling with SSE events

**3. Implementation Documentation**
Created comprehensive STREAMING_IMPLEMENTATION.md (200+ lines):
- Architecture overview
- Backend/frontend integration steps
- Code examples for hooks and components
- Testing procedures (curl + browser)
- Performance considerations (token buffering, memory management)
- Security (rate limiting, input validation)
- Troubleshooting guide
- Future enhancements

**Technical Details**
- Server-Sent Events (SSE) protocol
- Axios stream processing
- Buffer management for incomplete lines
- Delta content extraction from Azure responses
- Finish reason and usage metadata tracking

**Remaining Work (Frontend)**
- useStreamingChat hook implementation
- AIChatInterface streaming state management
- Token buffering for UI updates (50ms intervals)
- Streaming indicator and cursor animation
- Error recovery with fallback to non-streaming

**Impact**
- Perceived performance: Users see responses immediately
- Better UX: Token-by-token display feels more responsive
- Ready for production: SSE is well-supported across browsers
- Scalable: Can handle multiple concurrent streams

Files:
- src/modules/ai/ai.controller.ts: Added streaming endpoint
- src/modules/ai/streaming/streaming.service.ts: Core streaming logic
- docs/STREAMING_IMPLEMENTATION.md: Complete implementation guide

Next Steps:
1. Integrate StreamingService into AI module
2. Implement AIService.chatStream() method
3. Create frontend useStreamingChat hook
4. Update AIChatInterface with streaming UI

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

2025-10-03 22:24:53 +00:00

.duckversions

refactor: Consolidate settings page save buttons for better UX

2025-10-03 11:59:27 +00:00

ACCESSIBILITY_IMPLEMENTATION_PLAN.md

docs: Create comprehensive accessibility implementation plan

2025-10-02 20:50:16 +00:00

ACCESSIBILITY_PROGRESS.md

feat: Implement WCAG 2.1 AA accessibility foundation (Phase 1)

2025-10-02 21:35:45 +00:00

azure-openai-integration-summary.md

Add comprehensive .gitignore