Headline

Senior Data Scientist | GenAI & LLM Applications | ML & Recommendation Systems | IIM Bangalore

About

My foundation in data science is classical - Python, SAS, statistical modelling, propensity models, segmentation, hypothesis testing. The kind of work where you had to actually understand what you were doing because there were no pre-trained models to bail one out. That grounding still shapes how I think about machine learning problems, even when the tools look very different now. More recently, I've built an explainable AI coaching system (XGBoost + SHAP) across 55 sales KPIs - the goal was never just accuracy, but giving managers something they could actually act on the next morning. That one became a US patent application and a Dell Dyuti Best Project. Then a GraphSAGE recommender trained on 3.6M+ account-product relationships. Then LLM-powered reporting pipelines, hybrid RAG systems, agentic account planning workflows that compressed hours of research into under a minute. 10+ years of machine learning across fraud detection, loyalty analytics, e-commerce, media, and enterprise sales.

Location

India

AI stack

cursor

opencode

claudecode

Work experience

Advisor, Data Science at Dell Technologies, Feb 2022 - Present · 4 yrs 7 mos

Senior Data Scientist at Apexon (Client : Dell Technologies), Nov 2020 - Feb 2022 · 1 yr 4 mos, Developed quarterly propensity models at region × LOB granularity to score enterprise accounts by purchase likelihood. Outputs operationalised into sales targeting workflows with ~83% rep actioning rate on flagged accounts.

Lead Data Scientist at Tech Mahindra, May 2018 - Nov 2020 · 2 yrs 7 mos, Bengaluru, Karnataka, India, • Built a supervised audio segmentation pipeline to identify ad boundaries in TV broadcast streams, reducing manual review effort by ~45%. • Developed NLP-based text classification models for telecom support ticket routing and issue categorisation.

Program Manager, Data at Voonik, Sep 2016 - Feb 2018 · 1 yr 6 mos, Bengaluru Area, India, • Built a classification model to predict order cancellations - reduced cancellation rates by 23% through proactive intervention workflows. • Applied outlier detection to identify coupon abuse and return fraud patterns, reducing return rates by 19%.

Senior Business Analyst at PAYBACK, Apr 2014 - Sep 2016 · 2 yrs 6 mos, Bangalore, • Led campaign targeting and performance analytics for Future Group and other retail partners. Designed audience selection logic and measured post-campaign response across CRM channels. • Built transaction-based clustering models to segment retail stores nationwide, identifying high-performing store patterns for targeted marketing.

Education

Executive Programme in Business Analytics and Intelligence -8, Indian Institute of Management Bangalore

Bachelor’s Degree, Commerce, Delhi University