Jaskamal Kainth Software Engineer, Samsung

About

I build search systems that scale. For 8+ years I’ve worked on large-scale, intelligent search - from query preprocessing and retrieval to ranking, reranking, and weaving in NLP, AI/ML, and Gen AI.

I lead a search team that owns backend microservices, Search & Logging APIs, and end-to-end relevance. We took the platform from a single site to 55+ global markets while keeping it highly available and steadily improving relevance. My foundation is a dual degree in Mathematics and Computer Science (BITS Pilani), 300+ hours of DSA training and live problem-solving, and ACM-ICPC and competitive programming. I care about continuous learning and pushing AI-powered search forward.

Tech stack: AWS OpenSearch/Elasticsearch (BM25, k-NN with HNSW/Faiss), Spring-based microservices, FastAPI/Flask services, Redis, AWS infrastructure, PyTorch-powered ML, LLM inference frameworks (vLLM, TGI, Ollama) and agentic AI systems (MCP servers/tools), RESTful APIs, and strong fundamentals in data structures & algorithms.

Education

Birla Institute of Technology and Science, Pilani (BITS Pilani)
Dual Degree: B.E. (Computer Science) & M.Sc. (Mathematics), 2013 – 2018

Experience

Chief Software Engineer
2018 – Present
  • Lead engineering of Samsung's global AI-powered product search platform
  • Designed end-to-end search pipeline: query understanding → retrieval → ranking → reranking
  • Built scalable microservices and APIs powering search across 55+ global markets
  • Integrated NLP, ML, and Generative AI to improve search relevance and product discovery
  • Ensured high availability, performance, and observability for global search systems

Work & Writing

Academic work, technical notes, and interactive demos.

Pixelate Your Image

Upload any image and transform it into pixel art with character overlays.

ANN HNSW Simulation

Visualization of Hierarchical Navigable Small World for approximate nearest neighbor search.

Random Bounce Pattern (Dots)

Interactive particle simulation with bouncing dots.

Bloom Filters

Space-efficient probabilistic data structures; implementation and false-positive analysis.

7 Solutions to 1 Problem

Multiple approaches: triplets A[i] < B[j] < C[k].

Fast Fourier Transformation (FFT)

Multiply polynomials in O(n log n); applications for large number multiplication.

Elasticsearch Notes

Concepts and best practices for Elasticsearch.

C++ Notes

Programming language concepts and techniques.

Segment Tree Problems

Curated list of 20 problems, Easy to Hard.

String Algorithm Problems

Z-Algorithm, Trie, Suffix Array, Aho–Corasick, Hashing.

Disjoint Set Union (DSU) Problems

Problems using the DSU data structure.

Graph Theory Notes

Concepts and problem-solving techniques.

Beyond Code

Philosophy, language, music.

Philosophy Notes

Great thinkers of the last century - what we do, why we do it.

한국어 학습 여정

Korean learning journey - Hangul to grammar.

Music

Sonic landscapes.