ProjectJuly 12, 2026
Product Recommendation Engine
A recommendation system for suggesting relevant products to users — client work at IWConnect.
- Category
- AI & RAG
- Tech stack
- Recommendations
- Role
- Prototype — design & build
Content
A prototype recommendation system for surfacing relevant products from a large catalog. It explores the trade-off between personalisation quality and the cold-start problem of new users and new products.
Problem
In a large catalog, users cannot find relevant products without personalized suggestions.
Limitations
Prototype-stage: evaluated offline on test data, not against live user behaviour.
Next steps
A production version would need online evaluation, cold-start handling and a feedback loop from real interactions.
Outcome
Working prototype that produced personalised suggestions over a test catalog; deliberately kept at prototype stage.