About
Data scientist with 6+ years building production-grade ML systems like low-latency recommenders, contextual bandits, and LLM-powered products that have driven multi-million dollar revenue and measurable engagement gains.
What I'm looking for
Open to remote Data Science, AI PM, Product Management, and Project Management roles.
Experience
Senior Machine Learning Scientist
Dream11
Jul 2021 – Present
● Designed and implemented a forecasting model for dynamic supply, optimizing opportunity loss, driving 0.01% (~$6.3M) in incremental revenue
● Developed and deployed a machine learning model to predict the peak concurrency of ~16Mn users, enhancing system efficiency and ensuring 100% up time.
● Built and deployed a real-time push notification personalization system using affinity and CTR models with contextual bandits, driving 1.5-2% uplift in top-funnel metrics and 0.6% DAU growth while eliminating manual intervention and saving 100+ hours/month.
● Built and scaled real-time, low-latency live stream recommendation systems using ranking models and contextual signals, driving +2.34% CTR and +1.86% watch time uplift
● Developed a GenAI voice companion for sports watchalongs using an LLM-powered agent architecture. Led persona-driven prompt engineering and built evaluation frameworks for response quality. Implemented modular prompt orchestration (dynamic persona, tone, and context assembly) and integrated real-time event signals via event-driven pipelines
● Conducted multiple A/B experiments, systematically testing various business notions and leveraging insights to achieve revenue boost by 0.02%.
ForecastingContextual banditsLLM-powered agents
Business Technology Solutions Associate Intern
ZS Associates
Jan 2021 – Jun 2021
● Designed data pipelines to unify datasets and generate insights for key business questions, collaborating with cross-functional teams to define requirements and business rules using PostgreSQL, AWS, Redshift, and Python.
PostgreSQLAWSPython
Researcher
PathCheck, MIT Media Labs
Oct 2020 – Dec 2020
● Researched alongside a diverse team to study about COVID-19 Testing & Vaccine methodologies, challenges & solutions to publish research papers and build open-source applications for vaccine initiatives.
ResearchOpen-source applicationsCOVID-19 methodologies
Data Science Intern
Merkle Sokrati
Aug 2020 – Dec 2020
● Built an end to end propensity model with an accuracy of 84% for mid funnel optimization and deployed it on GCP.
Propensity modelingGCPMid funnel optimization
● Successfully built an audience of 100k+ with $0 spent in infrastructure & marketing, partnered with 100+ companies like Byjus, Great Learning, scaling to 150k+ MAV, generating 50mn+ impressions, with an engagement rate of 11.29%
● Overlooked the product, including market analysis, user interviews, creating user personas & roadmaps, building websites. Set up GTM strategy spotting trends & patterns, setting goals, distribution channels, pricing & branding
Product managementMarket analysisUser research
Education
Georgia Institute of Technology
Masters of Science · Computer Science
2024
University of Mumbai
Bachelor of Technology
2017 – 2021
Skills
GitPostgreSQLGCPAWSSQLPySparkPythonLow Latency SystemsFeature EngineeringA/B testing frameworksModel servingReal-time pipelinesPrompt EngineeringLLMsForecastingRanking ModelsRecommendation Systems
Languages
English (Native or bilingual proficiency)
