AI & Machine LearningTechnologyPlatinum

Design personalized recommendation systems for large user bases.

Recommendation System Designer

Collaborative Filtering, A/B Testing, Real-Time Serving

intermediatev6.0

Best for

▸Design two-tower neural collaborative filtering models for e-commerce product recommendations
▸Build cold-start solutions for new users with content-based hybrid approaches
▸Implement real-time candidate generation and ranking pipelines for streaming platforms
▸Optimize recommendation diversity and mitigate filter bubbles in social media feeds

What you'll get

▸Multi-stage architecture diagram with ANN retrieval, neural ranking models, and specific technologies (FAISS, TensorFlow Serving)
▸Feature engineering strategy combining implicit signals, content features, and contextual data with embedding dimensions
▸Evaluation framework including offline metrics (NDCG, MAP) and online A/B testing methodology with business KPIs

Expects

Clear description of the recommendation domain, user-item interaction data availability, catalog size, latency requirements, and specific business objectives.

Returns

Complete system architecture with candidate generation strategy, ranking models, cold-start handling, evaluation metrics, and production deployment considerations.

What's inside

“You are a Recommendation System Designer. You design personalized recommendation systems end-to-end, from candidate generation through ranking, re-ranking, and A/B testing, across catalogs ranging from hundreds to billions of items. - **Architecture before algorithm.** You never select a model befor...”

Covers

What You Do DifferentlyMethodologyWatch For

Not designed for ↓

×General machine learning model training or hyperparameter tuning
×Building basic search engines or information retrieval systems
×Creating simple rule-based filtering systems
×Database design for storing user preferences

SupaScore

89.63▼

Research Quality (15%)

9.1

Prompt Engineering (25%)

Practical Utility (15%)

8.65

Completeness (10%)

9.4

User Satisfaction (20%)

8.95

Decision Usefulness (15%)

8.8

Evidence Policy

Standard: no explicit evidence policy.

recommendation-systemcollaborative-filteringcontent-basedcold-startembeddingtwo-towerpersonalizationab-testingrankingcandidate-generationmatrix-factorization

Research Foundation: 7 sources (5 paper, 1 official docs, 1 books)

This skill was developed through independent research and synthesis. SupaSkills is not affiliated with or endorsed by any cited author or organisation.