I'm a Product Data Scientist

I identify growth, optimize for real values, and help teams cut through the noise.

*This is where I'd normally write a long (boring) paragraph about my qualifications and passion, but that would drop your retention by 40%, so let's skip it.*

I recommend checking out my highlighted case studies. They summarize all my experience in quick and fun reads.

Highlighted Case Studies

Feature Evaluation · Launch Decisions ⏱️ 5 min read

When Engagement Lifts Mislead

Engagement spiked after a feed update, but signals conflicted. Used Propensity Matching to isolate selection bias.

Explore the Analysis →
Growth Optimization · Budget Allocation ⏱️ 6 min read

Optimize for Incremental LTV, Not ROI

ROI favored high-volume channels with weak retention. Media Mix Modeling + Uplift Modeling surfaced long-term LTV drivers.

Explore the Analysis →
Performance Analysis · Root Cause Diagnosis ⏱️ 11 min read

Was the Feed Update the Culprit?

Engagement dropped post-launch, initially blamed on feature. Agentic Root Cause Analysis + Difference in Differences showed seasonality and user-mix shift.

Explore the Analysis →

Leadership & Impact

Built analytics products from zero to scale

Founding team member for multiple analytics products, shaping data foundations, metrics, and decision workflows from day one.

Scaled data-backed product operations globally

Led and scaled core product operations across 35 cities and 3 countries, enabling consistent execution, measurement, and sustained customer retention growth.

Research leadership with real-world focus

Lead author on 4+ peer-reviewed research papers spanning predictive maintenance, NLP automation, and risk modeling, bridging applied research and production systems.

Global hackathon winner

First place in a global AI hackathon (1,300+ participants across 19 countries), recognized for building an end-to-end, production-minded AI solution under constraints.

Background

Professional Experience

Data Science Researcher

Julius Baer

Jun 2025 – Sept 2025
Gen AI Personalization Experimentation
  • Performed causal and exploratory analysis on customer engagement data, driving a 25% lift in content engagement.
  • Developed LangGraph and Vertex AI personalization chatbot; automated content selection and cut curation time by 35%.
  • Established experimentation metrics framework, enhancing A/B test accuracy, guardrails, and reproducibility across teams.

Data Scientist Apprentice

TD Bank

Jan 2025 – May 2025
Customer Analytics Fraud Detection Credit Scoring
  • Executed causal inference and A/B testing to evaluate fraud-detection strategies on 100M+ transactions.
  • Engineered XGBoost in Spark for credit scoring, raising PR-AUC from 0.20 to 0.67 (235% lift) with SHAP explainability.
  • Collaborated with senior management and risk teams to inform segmentation, fraud reduction, and credit strategy.

Quantitative Research Apprentice

Ask2AI

Jan 2025 – May 2025
Acquisition Analytics Growth Optimization
  • Implemented uplift modeling and experimentation to isolate causal impact on acquisition, improving ROI attribution by 22%.
  • Constructed regression and MILP optimization models for acquisition allocation, raising customer lifetime value by 12%.
  • Delivered forecasting-driven insights to guide growth strategy, increasing acquisition rates and decreasing churn by 25%.

Teaching Assistant

Columbia University

Sept 2024 – Jan 2025
Algorithms to Data Science
  • Mentored 250+ MS students, bridging ML algorithms with product-focused analytical reasoning and experimentation frameworks.
  • Designed case studies and assignments emphasizing causal inference, A/B test design, and data-driven decision-making.

Data Scientist

Navin Fluorine International Limited

May 2023 – Nov 2023
Predictive Maintenance Operational Analytics
  • Built predictive maintenance models reducing downtime cost by 35%; automated KPI pipelines cutting latency by 40%.
  • Deployed anomaly-detection systems surfacing high-risk events earlier, improving recall by 15%.

Founding IT Intern

E-Revbay Pvt. Ltd.

Dec 2021 – May 2022
Acquisition Analytics Retention Analytics
  • Constructed customer acquisition prediction model using Equifax data, improving campaign precision by 50%.
  • Automated dashboards and root-cause pipelines, reducing intervention time by 25.

Education

M.S. Data Science

Columbia University · New York, NY

Sep 2024 – Dec 2025
GPA 3.7/4.0
  • Relevant Courses: Agentic AI, Fintech, and Data Economy (PhD elective), Big Data Analytics, Applied Machine Learning, Statistical Inference & Modeling, Algorithms for Data Science (TA), DS Applications in Insurance & Banking

B.Tech Honors (Computer Engineering, Data Science/Analytics)

NMIMS University · Mumbai, India

Jun 2020 – May 2024
GPA 3.9/4.0
  • Relevant Courses: Deep Learning; NLP; Reinforcement Learning; Computer Vision

Research Publications

Predictive maintenance for metro systems

Google Scholar · Lead Author

Demonstrated a sensor-driven pipeline that predicts equipment failures so operations teams can schedule targeted maintenance and reduce downtime.

Diabetes detection optimized for recall

Google Scholar · Lead Author

Prioritized recall in model design to reduce missed diagnoses, improving early detection reliability for clinical use.

NLP to SQL for mobile learning

Google Scholar

Built an NLP-to-SQL interface that lets non-technical users query student and CSV data directly from mobile devices.

Semi-supervised disease prediction

Google Scholar

Applied semi-supervised methods to leverage unlabeled clinical data and improve prediction robustness in ambiguous diagnostic cases.

Contact

Always happy to talk product, data, or research. I reply within 24 hours (EST).

Email: abhitayshinde@gmail.com LinkedIn GitHub