Department of Statistics of St. Xavier’s University, Kolkata (SXUK) organized an online guest lecture titled “Retrieval at Scale: System Design for AI-Driven Search” on February 23, 2026 at the university campus.

Noted researcher Dr. Shankar Pal was the resource person. He delivered an in-depth and comprehensive presentation on the architecture of large-scale search systems and their integration with modern artificial intelligence technologies. He emphasized the growing need for intelligent, efficient, and scalable search mechanisms in contemporary digital applications.

Dr. Pal discussed the growing importance of hybrid search and Retrieval-Augmented Generation (RAG), highlighting key challenges in large-scale retrieval systems such as latency reduction, distributed storage efficiency, index consistency, data freshness, relevance optimization, and cost management.

Dr. Pal also outlined the technical foundations of modern search infrastructures, explaining inverted indexes, vector databases, and the steps involved in building vector indexes—including embedding generation, normalization, sharding, and graph-based construction.

A major focus was the Hierarchical Navigable Small World (HNSW) algorithm, which enables fast and accurate nearest-neighbor retrieval. The session saw enthusiastic student participation and was widely appreciated for its clarity and depth, offering valuable insights into AI-driven search technologies and their practical applications.

Dr. Soma Debnath