Open to opportunities

Madhura Dasgupta

@madhuradasgupta

Message

Data scientist and economist delivering predictive models and actionable analytics for business impact.

India

Message

What I'm looking for

I seek roles applying advanced ML and econometric methods to real-world forecasting, recommendation, or credit-risk problems, with collaborative teams and impact focus.

I am a data scientist and economist with a Ph.D. and 5+ years of experience building statistical and machine learning solutions across healthcare, finance, and education.

I have developed forecasting, recommendation, and credit-scoring systems that produced measurable business value, including high-accuracy XGBoost models and ensemble forecasting approaches.

I design end-to-end workflows in Python, implement robust evaluation frameworks, and tune models for generalization, achieving strong precision, recall, and AUC results in applied projects.

I also have academic experience teaching and mentoring, creating curricula that combine economic theory and data analytics, and publishing interpretable machine-learning research on enterprise borrowing.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Scientist

Current

Thoucentric Technology Private Limited

Jun 2025 - Present (8 months)

Lead forecasting initiatives to predict market demand and consumption patterns and develop elasticity models to analyze price sensitivity and product switching, automating data processing and model training workflows using Python.

Python Statistical Modeling Pandas

Assistant Professor

Indian Institute of Management Bodh Gaya

Jul 2024 - Mar 2025 (8 months)

Taught postgraduate Microeconomics and Macroeconomics, developed curriculum integrating economic theory with practical data-analytic applications, and mentored students on research projects combining economics and analytics.

Curriculum Development Mentoring Microeconomics Data Analytics Academic Assessment

Associate (AI team)

ConveGenius Digital India Services Pvt Ltd

Dec 2022 - Oct 2023 (10 months)

Designed intra-bot recommendation systems including hybrid collaborative filtering and SVD-based models with popularity fallback, achieving ~75% precision and ~70% recall and MAE ≈ 0.75 on a 15-point scale.

Recommender Systems Collaborative Filtering Python Model Evaluation Precision Recall MAE Cold Start Strategies

Postdoctoral Researcher

Indian Institute of Management Ahmedabad

Jan 2022 - Nov 2022 (10 months)

Developed an ensemble forecasting framework for monthly sales prediction of bronchodilators using time series decomposition and Random Forest on residuals, improving predictive accuracy on pharmaceutical sales data.

Time Series Analysis Random Forest Python Seasonal Decomposition

Associate Fellow

National Council of Applied Economic Research

Sep 2019 - Dec 2021 (2 years 3 months)

Built XGBoost predictive models (94% AUC) to analyze determinants of Self-Help Group borrowing, conducted permutation feature analysis and counterfactual simulations to identify drivers of credit demand.

XGBoost Feature Importance Python AUC Evaluation

Education

Degrees, certifications, and relevant coursework

University at Buffalo, State University of New York

Doctor of Philosophy, Economics

2014 - 2019

Activities and societies: Teaching assistantship with full tuition scholarship; research and publications on applied machine learning for economic questions.

Completed a Ph.D. in Economics with research combining statistical modeling and machine learning applied to economic and healthcare datasets.