About the role
Responsibilities:
- Search Infrastructure Ownership: Own the search platform end-to-end, including: Indexing and data modelling, Query understanding and interpretation, Retrieval mechanisms, ML-powered ranking systems, Result serving infrastructure.
- Real-Time Indexing and Data Pipelines: Develop and maintain real-time indexing pipelines to synchronise: Inventory updates, Pricing changes, Product catalogue modifications from dark stores. Ensure search indices remain fresh and accurate to support quick-commerce operations.
- Ranking and Personalisation: Design and optimise ranking algorithms that incorporate: Product relevance, User personalisation signals, Inventory availability, Business and merchandising rules. Continuously improve ranking quality to maximise engagement and conversion.
- Discovery Experience: Build and enhance discovery capabilities beyond traditional search, including Personalized feeds, Category ranking systems, Similar product recommendations, and "Complete the Look" experiences
- Experimentation and Performance Optimisation: Implement A/B testing frameworks to evaluate and improve search experiences. Define, monitor, and optimise key search quality metrics, including MRR (Mean Reciprocal Rank), NDCG (Normalised Discounted Cumulative Gain), CTR (Click-Through Rate), and Conversion Rate. Optimise systems to maintain P95 latency below 200 milliseconds.
Requirements:
- Bachelor's or Master's degree in Computer Science, Engineering, or a related field.
- 5-7 years of experience building production-grade Search/Information Retrieval (IR) systems, preferably in e-commerce environments.
- Strong expertise in Elasticsearch/OpenSearch, including: Index design, Custom analyzers, Query DSL, and Performance tuning and optimisation.
- Hands-on experience with Machine Learning ranking techniques, such as Learning-to-Rank (LTR), XGBoost, LightGBM, and feature engineering.
- Experience with embedding-based retrieval systems, including: FAISS and HNSW.
- Strong programming proficiency in Python.
- Solid understanding of Natural Language Processing (NLP) concepts for query understanding, including: Tokenisation, Entity extraction, and Intent classification.
- Experience working with real-time indexing technologies, such as Kafka and CDC (Change Data Capture).
- Familiarity with search quality experimentation and evaluation frameworks.