About Me

With 15+ years of experience as a Strategic Data Science Leader, I specialize in leveraging AI, Machine Learning, and advanced analytics to drive business impact for Fortune 500 companies. My expertise spans forecasting, product recommendations, inventory optimization, marketing analytics, and Large Language Models (LLMs), with a proven track record of delivering high-performing solutions across industries like Cards & Payments and Healthcare Manufacturing & Supply Chain.

Education

M.S in Analytics, Georgia Institute of Technology

B.S in Electronics, Sikkim Manipal Institute of Technology

Work Experience

Principal Data Scientist @ Stryker Corp (March 2022 - Present)

• Partnering with senior executives to deeply understand business objectives and challenges, translating them into actionable analytic product features that drive measurable business outcomes.
• Building Implementation Stategy for Large Language Models (LLMs) and Agent Frameworks, setting the foundation for future-proofing LLM capabilities and driving innovation across the organization.
• Developing effective analytics strategies for a major marketing campaign, optimizing decision-making and measuring impact to drive customer engagement and ROI.
• Implementing accurate demand forecasting models, contributing to better resource allocation and strategic planning, enhancing operational efficiency.
• Led the development and deployment of GPT-based solutions for company policies, as well as customer recommendation and product association models across multiple divisions.
• Delivered comprehensive analysis on the effectiveness of major marketing initiatives, providing actionable insights and return on investment.

Lead Data Scientist @ Medline Industries (_March 2019 - March 2022)

• Established and led the Data Science team, constructing Machine Learning and Cloud Infrastructure, and expanding it to a 4-member team.
• Developed a smart demand forecasting system for inventory management, reducing forecasting error from 67\% to 32\% and generating yearly savings of over $12 million in excess inventory.
• Engineered a smart warehouse inventory audit application utilizing a Random Forest model, leading to weekly inventory adjustments worth $50,000+ across 5 warehouses. Currently scaling the application for 54 branches, with an estimated savings projection of $4 million.
• Implemented a real-time pricing enquiry application utilized by 40+ salesmen, resulting in the conversion of 19 customers from competitors with a 1-year contract valued at $10 million.

Sr Associate/Lead Data Scientist @ Cognizant (_Dec 2007 - Mar 2019)

• Established and led a 7-member data science team that successfully delivered end-to-end data science applications.
• Developed the Transaction Settlement Analytics tool, utilizing Decision Tree Regressor to predict merchant funding and saving $2 million in SLA fines over a span of 2 years.
• Orchestrated the architecture and implementation of the Google and Samsung Pay application for a client with a 40-member team over 8 months, resulting in an annual combined revenue of over $70 million.
• Constructed a real-time credit line increase decision engine application using Random Forest Regressor and Classifier, facilitating instant approval decisions and amounts, leading to significant time savings compared to the previous 3-5 business day processing time.

Projects

Ryanair Passenger Reviews Analysis

In the rapidly growing aviation industry, analyzing customer feedback is critical for improving service quality and customer experience. This project focuses on leveraging Natural Language Processing (NLP) techniques to analyze Ryanair passenger reviews from 2012 to 2024. By employing sentiment analysis, topic modeling, and root cause analysis, the study identifies key areas of satisfaction and dissatisfaction, highlights operational challenges, and pinpoints routes with recurring issues. A custom machine learning-based sentiment classifier is developed to automate sentiment categorization, and innovative graph network visualizations are used to trace negative sentiments back to specific routes. This analysis aims to provide actionable insights to enhance Ryanair’s operational and customer service strategies.

Report

Project

Amazon Product Bundling and Recommendation

The objective of this project is to conduct a comprehensive analysis of Amazon sales data to gain valuable insights into sales trends, customer purchase behavior, and other factors influencing profitability. Leveraging various data mining techniques such as exploratory data analysis, clustering, association mining, and predictive modeling, we aim to extract meaningful patterns and relationships to build a product bundling strategy and recommend focused products to customers. By delving into multifaceted dimensions such as product categories, geographical distributions, and sales performance metrics, this analysis will furnish actionable insights to refine sales strategies and augment overall profitability.

Report

Project

IntelliFraud: Bank Account Fraud Detection

IntelliFraud is an innovative online tool designed to enhance the detection of fraudulent bank account applications. It features an intuitive dashboard for exploring statistical data, dynamic graph network analysis to visualize and analyze transaction flows, and advanced machine learning techniques like Voting and Stacking Classifiers for fraud detection. The tool provides four key functionalities: exploratory data analysis (EDA) across customizable datasets, fraud network analysis for transaction visualization, model evaluation to assess classification metrics and identify optimal models, and an inference interface to analyze fraud cases and understand model decisions using SHAP and ELI5. IntelliFraud empowers users with actionable insights to improve fraud detection strategies effectively.

Report

Project

Case Study

Optimizing cancer patient care with advanced analytics The Oscar Lambret Center, a premier cancer treatment and research institution in Lille, France, embarked on a research initiative to establish personalized care pathways for cancer patients. These pathways are designed to guide physicians in delivering tailored and consistent care for improved patient outcomes. The project aligns with the center’s mission of providing high-quality care and fostering multidisciplinary collaboration. By leveraging advanced analytics, the center aimed to consolidate siloed data and create dynamic, actionable insights to enhance care quality and operational efficiency. Hospital faced challenges in delivering personalized care pathways for cancer patients wanted to improve patient outcomes by establishing benchmarks for standard treatment times to prevent care delays and Optimizing workforce allocation and resource utilization for multidisciplinary cancer care. Success story has been referred from : Oscar Lambret Center Success Story | Report