Product Backlog
Active backlog of features, improvements, and technical debt for Vows Social AI.
Last Updated: 2025-10-11
Current Sprint (Week 1)
Sprint Goal: Set up free tier infrastructure and foundational services
In Progress
- Qdrant Cloud Setup (P0)
- Create 1GB free tier cluster
- Configure collections for content/user embeddings
- Test vector search performance
-
Owner: TBD | Est: 4 hours
-
Supabase Database (P0)
- Deploy PostgreSQL on free tier
- Run initial schema migrations
- Set up connection pooling
-
Owner: TBD | Est: 4 hours
-
Cloudflare Workers Config (P0)
- Configure Workers environment
- Set up Durable Objects
- Configure KV store
- Set up secrets management
-
Owner: TBD | Est: 6 hours
-
Fly.io ML Service (P0)
- Deploy Python inference service
- Configure 3 free VMs
- Test embedding generation
- Owner: TBD | Est: 6 hours
Upcoming (Week 2-4)
Phase 1: Foundation Model
- Sentence-BERT Integration (P0)
- Load
all-MiniLM-L6-v2model - Implement content embedding
- Deploy to Fly.io
-
Owner: TBD | Est: 8 hours
-
User Embedding Logic (P0)
- Weighted aggregation implementation
- Recency weighting
- Cold start handling
-
Owner: TBD | Est: 12 hours
-
Thompson Sampling (P0)
- Beta-Bernoulli implementation
- Durable Object integration
- Reward update logic
-
Owner: TBD | Est: 16 hours
-
Feed Generation API (P0)
/api/feed/:userIdendpoint- Vector similarity ranking
- Caching strategy
- Owner: TBD | Est: 12 hours
Backlog (Phase 2+)
High Priority
Multi-Agent System
- Discovery Agent (P1)
- Instagram API integration
- Curator account management
- Quality filtering with Gemini
-
Est: 2 weeks
-
Quality Guardian (P1)
- Multi-modal quality scoring
- GPT-4V integration
- Tiered evaluation pipeline
-
Est: 1 week
-
Personal Archivist (P1)
- Planning phase inference
- Style evolution tracking
- Timing intelligence
-
Est: 1 week
-
MAGRPO Framework (P1)
- Group sampling implementation
- Advantage computation
- Simultaneous policy updates
- Est: 2 weeks
Infrastructure
- Custom Foundation Model (P1)
- Transformer architecture design
- Training pipeline (Colab)
- A/B testing framework
-
Est: 3 weeks
-
Monitoring & Observability (P1)
- Metrics dashboards (Grafana)
- Error tracking (Sentry)
- Performance monitoring
- Est: 1 week
Medium Priority
Features
- Serendipity Engine (P2)
- Diversity measurement
- Cross-regional discovery
- Style exploration
-
Est: 1 week
-
Engagement Forecaster (P2)
- 200+ context signals
- Welcome probability prediction
- Send-time optimization
-
Est: 2 weeks
-
Search Functionality (P2)
- Semantic search API
- Filter by vendor type, region, style
-
Est: 1 week
-
Collections (P2)
- Save content to collections
- Share collections
- Collection management UI
- Est: 1 week
Optimization
- Edge Caching (P2)
- Multi-tier cache strategy
- Cache invalidation logic
- Performance benchmarking
-
Est: 3 days
-
Database Optimization (P2)
- Query optimization
- Index tuning
- Connection pooling
-
Est: 3 days
-
Batch Processing (P2)
- Batch LLM calls
- Batch vector searches
- Cost optimization
- Est: 3 days
Low Priority
Nice to Have
- Mobile App (P3)
- Flutter implementation
- Push notifications
- Offline mode
-
Est: 6 weeks
-
Vendor Analytics Dashboard (P3)
- Profile views tracking
- Inquiry analytics
- Engagement metrics
-
Est: 2 weeks
-
Admin Dashboard (P3)
- User management
- Content moderation
- System metrics
-
Est: 2 weeks
-
Social Features (P3)
- Follow vendors
- Share with partner
- Comment on posts
- Est: 2 weeks
Technical Debt
High Priority Debt
- Error Handling (P1)
- Comprehensive error handling across all services
- Graceful degradation
- Circuit breakers
-
Est: 1 week
-
Security Audit (P1)
- API authentication/authorization
- Input validation
- Rate limiting
- Secrets management review
-
Est: 3 days
-
Testing Coverage (P1)
- Unit tests for all agents
- Integration tests
- E2E tests
- Target: 80% coverage
- Est: 2 weeks
Medium Priority Debt
- Documentation (P2)
- API documentation (OpenAPI)
- Deployment runbooks
- Architecture diagrams
-
Est: 1 week
-
Code Quality (P2)
- TypeScript strict mode
- ESLint configuration
- Python type hints
-
Est: 3 days
-
Performance Testing (P2)
- Load testing (10K QPS)
- Latency benchmarks
- Memory profiling
- Est: 3 days
Completed ✅
Week 0 (Documentation)
- Architecture Documentation
- PRD finalized
- System architecture designed
- ADRs documented
-
RFC-0001 created
-
Documentation Site
- MkDocs setup
- Getting started guides
- Component documentation
- Implementation plans
Feature Requests
Track user-requested features here
From Beta Users
- None yet (pre-launch)
From Team
- Vendor verification - Allow vendors to claim profiles
- Budget filtering - Filter content by price range
- Timeline integration - Sync with Google Calendar
- Email digests - Weekly vendor recommendations via email
Prioritization Framework
P0 (Critical): Blocks launch, must do now P1 (High): Important for Phase 1/2 success P2 (Medium): Valuable but can wait P3 (Low): Nice to have, future consideration
Effort Estimation: - 1-4 hours: Tiny - 4-8 hours: Small - 1-2 days: Medium - 3-5 days: Large - 1-2 weeks: Very Large - 2+ weeks: Epic (break down)
How to Add Items
- Create new item with clear description
- Assign priority (P0-P3)
- Estimate effort
- Tag with relevant labels
- Assign owner if known
- Link to related issues/PRs
Example:
- [ ] **Feature Name** (P1)
- Clear description of what needs to be done
- Acceptance criteria
- Owner: @username | Est: X hours
- Related: #123, #456
Sprint Planning
Velocity: TBD (will track after first sprint)
Sprint Cadence: 1 week sprints
Planning Schedule: - Monday: Sprint planning - Daily: Standups - Friday: Sprint review & retro