Paras Lehana
Senior Lead Engineer at IndiaMART · Full‑stack & AI enthusiast
lehana
Noida, Uttar Pradesh, India
Joined October 2025
Network
1.1K connectionsCGRKCB
AKVTAN
DIRGSS
RKCGAL
PSSSRR
YYGEVK
SSRKVA
Summary
Product-search and Auto‑Suggest engineer at IndiaMART: Paras has focused much of his professional work on product search and Auto‑Suggest systems, contributing as an individual contributor and then as a lead to production search features, performance tuning and AI/ML POCs. He is active in internal innovation (IndiaMART hackathons) and has helped ship and prototype agentic/AI features for the company. atlassian+2
Active open‑source developer with data/ML projects: He publishes code and tools on GitHub (notably chess-stats-ml and several utility projects), demonstrating practical skills in Python, ML/data workflows, and full‑stack development. github+1
Technical writer and blogger on systems, ML and broader tech topics: Maintains a technical blog / Medium presence where he writes about systems (memory/GC/performance), AI/ML topics and broader technology and societal issues. lehana+1
Competitive chess and sports background informs analytical thinking: Paras has co‑authored academic work on chess prediction and lists chess and badminton among his activities; this sporting background and his chess research reflect an analytical, data‑driven approach to problem solving. academia+1
Work
Education
Projects
Writing
Python PyPIM breaks the Memory Wall Problem
November 1, 2024Medium / personal blog post exploring an approach to reduce data-transfer bottlenecks by processing inside memory — an example of his technical writing on systems and performance.
Statistical Analysis on Result Prediction in Chess
January 1, 2018Academic paper describing a system that annotates tens of thousands of chess half‑moves with Stockfish evaluations and uses machine‑learning (multivariate linear regression and classification) to predict game outcomes and recommend moves.