Headline

Uber | Ex Playgames24x7 (PPO)| PGDBA (Institute Rank - 2, IIM Cal, ISI Kol, IIT Kgp) | IIT Bombay | Hindu College (Delhi University)

About

I am a Data Scientist with 4+ years of experience applying machine learning and analytics to solve high-impact business problems. My work focuses on personalization, predictive modeling, and data-driven strategies that drive revenue growth, efficiency, and user engagement. At Games24x7, I built large-scale targeting and personalization frameworks that improved conversions, reduced acquisition costs, and delivered multi-crore revenue impact. Previously at Abbott Nutrition, I led sales force effectiveness analytics and created automated dashboards that enhanced field force productivity and commercial excellence. With a strong academic foundation in mathematics and business analytics (PGDBA – IIM Calcutta/ISI Kolkata/IIT Kharagpur, MSc – IIT Bombay), I bring a mix of technical expertise and business insight, with a passion for building scalable AI/ML solutions.

Location

Bengaluru

AI stack

claude

cursor

hugging face

langchain

notebooklm

openai

Work experience

Scientist 2 tech at Uber, Dec 2025 - Present · 9 mos, Bengaluru, Karnataka, India, • Enhancing earners and user experience via improved trip completion guardrails. • Merchant notes summarization

Data Scientist at Games24x7, Apr 2025 - Oct 2025 · 7 mos, Player latent representation • Implemented BERT-inspired masked language modelling (MLM) framework to learn latent player embeddings from sequential gameplay events  automatic representn learning w/o manual feature engineering. • Trained transformer-based embeddings capturing contextual relationships among players, achieved accuracy of ~95% | Integrated XGBoost models for downstream tasks such as churn prediction & conversion modelling. • Validated the transferability of player embeddings, demonstrating that once trained via MLM, they can be reused across multiple predictive tasks, improving ROC-AUC from 0.67 (XGBoost) to 0.85 (with BERT-MLM) First deposit personalization • Built a first-deposit personalization framework that recommends deposit tiles by combining an XGBoost-based CP model with a location-aware nearest neighbour distribution of user deposits. • Optimized NN search using ANN, ↓ing computation time from 7–8 sec to <0.05 sec/user with ~95% accuracy. • On A/B testing it gave us a lift of ~6-7%

Data Science Intern at Games24x7, Oct 2024 - Apr 2025 · 7 mos, PPO offered my11circle customer acquisition, customer retargeting: • Optimized marketing spends by better targeting | Designed & deployed a large-scale user re-targeting pipeline • Engineered 50+ behavioural, calendar & recency-based features & built XGBoost models achieving up to 0.80 ROC-AUC, also used predicted baseline conversion ratio (using prophet timeseries modelling) as an input var. • Delivered ~1.2 lakh additnal user conversions  ~₹7 crore additional revenue by improving targeting efficiency, reducing MCPA by ~60% & generating 4-6X lift in top 5% predicted cohort compared to baseline conversions.

Business Analyst at Quess Corp Limited, Sep 2017 - Oct 2018 · 1 y 1 mo, Mumbai, Maharashtra, India, client: Abbott Nutrition India Got the highest performance grade for consecutively two years • Led SFE analytics, tracking KPIs (HCP coverage, TCFA, FWD, missed calls, chemist call average, call frequency, specialty coverage, P/P, P/D, attrition, compliance, ESS/EPK score, CME/RTM in clinic camps) linking them to business performance (PCPM, growth, sales achievement, OPSC NOD improvement) for timely leadership interventions. • Helped prioritize high-potential HCPs, improving PCPM sales by building a regression model (Random forest) to predict expected Rx-growth/HCP based on historical call frequency, specialty coverage, of in-clinic camps, dietician assigned, CME participation, and prior prescription trends. • Saved cost on import expiry inventory by developing accurate time series demand forecasting model (FB Prophet model) using historical sales data. Built automated dashboards for sales performance, SFE KPIs, market share, prescription trends, saving approximately 0.5 day/month for 450+ FF. • Conducted market share, growth penetration analysis and participated in sales reviews field work; data liaison for senior management, rationalize discounting across hospitals and supported brand planning. • Improved data quality for 55000 HCPs 1500+ Hospitals NHs. Recommended cross-engagement, dual engagement allocation strategies (dieticians, TBEs, NHs) to maximize HCP coverage. • Recognized as a top performer (EE rating, top 5%) for 2 consecutive years for driving data-driven decision-making, FF productivity

Business Analyst at Alp Consulting Ltd., Jun 2016 - Aug 2017 · 1 y 2 mo, Mumbai, Maharashtra, India, client: Abbott Nutrition India

Education

PGDBA, Business analytics, Indian Institute of Management, Calcutta

PGDBA, Business analytics, Indian Institute of Technology, Kharagpur