Round 1: Technical screening — SQL (window functions, aggregations, cohort-style queries), classical ML fundamentals, statistics, and product sense/metrics questions. Standard DS screening format.
Round 2: Deep technical round with a senior IC. Covered ML concepts in depth, NLP/search systems, project deep-dives with trade-off justifications, and Blinkit-specific problem framing (demand forecasting, ETA prediction, search ranking). Expect to defend your design decisions end-to-end.
Round 3: Hiring manager round.
Preguntas de entrevista [1]
Pregunta 1
They gave me a metric debugging case something like "order delivery times have increased in a specific city, diagnose why."
Me postulé a través de un reclutador. El proceso tomó 1 día. Acudí a una entrevista en Blinkit (Gurgaon, Haryana) en feb 2025
Entrevista
Previous experience based questions and some general data science questions. Some of them were based on recommendation systems and classical ML like how to recommend top products, bagging vs boosting etc
Preguntas de entrevista [1]
Pregunta 1
- Difference between Bagging and Boosting and their workings.
- Precision and recall
- Word2Vec working
- Activation functions
- Factors you would consider while recommending products on the page.
- How to create basket of items which a user might buy together? eg for gym goers -> protein based items, dumbells etc
Me postulé en línea. El proceso tomó 2 días. Acudí a una entrevista en Blinkit
Entrevista
The interview process was very simple. It had 2 rounds. First round was technical and second round was behavioral interview. Technical round solely went on previous projects , technologies used for them and business logic applied behind these. There was a SQL and python question asked at the end.