Major project
Retail, E-commerce & Consumer Goods
Shopping assistant with sources
A RAG assistant over a store's catalogue and return policy that answers shoppers and links the products.
The brief
Shoppers ask questions a search box cannot answer: "a cotton kurta under ₹1,500 for office wear". Build an assistant over a product catalogue and the store's policies that recommends real products with links, answers policy questions with the clause, and never invents products. Good looks like: retrieval compared on 50+ shopper queries, product links on every recommendation, and refusals when nothing matches.
What you will build
- Load the catalogue and policies
- Build hybrid search with filters for price, size and category
- Add re-ranking and product links
- Write 50+ shopper queries with good answers
- Evaluate relevance and faithfulness
- Deploy the assistant
Free, open data to build it with
- Shopping Queries Dataset (ESCI) ↗ Amazon Science (GitHub)
- Flipkart products ↗ Kaggle
Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.