Social Media Algorithm Developer

We are seeking an experienced **Algorithm Developer** to design and implement a video recommendation algorithm for our emerging social media platform. This is a time-sensitive project with a 1-day delivery timeline.

## Project Scope
Design and deliver a functional recommendation algorithm that suggests relevant videos to users based on engagement patterns and content similarity. The algorithm must integrate seamlessly with our existing tech stack and be production-ready for initial deployment.

## Technology Stack
- **Backend Framework:
Node.js + Express.js
*Database:** MongoDB
- Frontend:** React.js
- Deployment:Self-hosted infrastructure (Docker preferred)

Core Requirements

Algorithm Specifications
1. Recommendation Strategy:Hybrid approach combining collaborative filtering and content-based filtering
2. Candidate Generation: Multi-source candidate retrieval (followed creators, similar content, trending videos)
3. Ranking Model: Scoring mechanism based on:
- User engagement history (watch time, likes, shares, comments)
- Video metadata (category, tags, duration, creation date)
- Completion rates and skip behavior
- Time-based decay (favor recent engagement)
- Content similarity embeddings

aDta IntegrationThe algorithm must work with existing MongoDB collections:
- *mUsers collection:* User profiles, preferences, viewing history
- Videos collection: Video metadata, embeddings, creator information
- Interactions collection: Watches, likes, shares, comments, skips
- Follows collection: User-to-creator relationships

API Endpoints
Implement REST API endpoints:
1. `GET /api/recommendations/:userId` - Fetch personalized feed (20-50 video recommendations)
2. `POST /api/interactions/:videoId` - Log user interactions (watch, like, skip)
3. `GET /api/trending` - Trending video recommendations
4. `POST /api/algorithm/retrain` - Trigger model retraining (optional admin endpoint)

Deliverables
1. Algorithm Implementation: Complete working code in Node.js/Express.js
2. MongoDB Queries: Optimized aggregation pipelines for data retrieval
3. API Integration:Fully functional REST endpoints ready for React.js frontend
4. Documentation: Code comments, API documentation, and implementation notes
5. **Performance Metrics:** Expected response times and recommendation quality metrics
6. **Docker Setup (Optional):** Containerized deployment configuration

## Technical Expectations

### Performance Requirements
- API response time: < 500ms per request
- Support concurrent users: Minimum 1000 simultaneous requests
- Daily retraining capability for model updates
- Scalable architecture for future growth

### Code Quality
- Clean, well-documented code
- Error handling and validation
- Logging for debugging and monitoring
- Unit tests for core functions

### Algorithm Efficiency
- Recommend videos within available context window
- Balance between recommendation diversity and personalization
- Avoid creating filter bubbles (include serendipitous recommendations)
- Optimize database queries to minimize latency

## Nice-to-Have Features
- A/B testing framework for algorithm variations
- Recommendation explainability (why this video was recommended)
- Cold-start handling for new users/videos
- Admin dashboard for monitoring algorithm performance
- Multi-language support for content categorization

Budget & Timeline
- Deadline: 1 day from project
- Preferred Candidates:** Developers with prior experience in recommendation systems or machine learning engineering

Ideal Candidate Profile
- 3+ years experience with Node.js and Express.js
- Proven MongoDB expertise with complex queries/aggregations
- Understanding of recommendation algorithms and machine learning concepts
- Experience with recommendation systems or content discovery algorithms
- Ability to deliver production-ready code under tight deadlines
- Strong problem-solving skills and communication abilities
- Experience with self-hosted deployment (Docker, Linux preferred)

How to Apply
- Portfolio or GitHub links demonstrating algorithm/ML work
- Brief overview of your approach to building this algorithm
- Availability confirmation for immediate start
- Technical references or past project examples

Project Communication
- Daily check-ins during development
- Real-time Slack/Discord communication
- Immediate feedback on deliverables
- Post-delivery support for integration assistance