Design Product Search Ranking (Basic)

Beginner 45 minAmazonFlipkartMeeshoMyntraWalmart

Design product search ranking for an e-commerce site: given a query like 'running shoes for men', return the most relevant products that users are likely to buy.

Functional requirements

Retrieve candidate products for a query
Rank them by relevance and purchase likelihood
Support filters (price, brand) and sorting
Handle misspellings and synonyms

Non-functional requirements

Results in under 200 ms
Fresh inventory and prices
Measurable business impact (CTR, conversion)

Scale

Users: 50M DAU
Traffic: ~20K searches/s at peak
Storage: 100M products

Hints · 0/3 used

Hint 1Sign in to use hints
Ranking 100M products per query is impossible. What is the first stage?
Hint 2Sign in to use hints
Retrieval (BM25 / embeddings) gives a few thousand candidates; a ranker orders them. What labels train the ranker?
Hint 3Sign in to use hints
Query understanding (spell-correct, synonyms, category prediction) → retrieval from inverted index plus vector search → learning-to-rank (gradient-boosted trees) on features like text match, price, ratings, CTR, and conversion → business rules; labels from clicks/purchases with position-bias correction; evaluate NDCG offline and A/B test CTR/conversion.

How you'll be scored

Requirements10
Data15
Model15
Serving15
Evaluation15
Monitoring10
Cost & latency10
Trade-offs10

What the interviewer expects in each area unlocks after your first submission.

Learn it from MIK

Solve it in the workspace

Write requirements and estimates, draw the architecture, and get your design scored. Your drafts save as you type and follow you across devices.