Open to opportunities

Karmabir Chakraborty

@karmabirchakraborty

Message

Data Scientist at LS Direct who connected ChatGPT to governed BigQuery data and improved campaign ROI by 25%.

United States

Message

At LS Direct, I connected ChatGPT to governed BigQuery tables, rules, and regulations through MCP. I also built customer behavior models across e-commerce clients, increasing campaign ROI by 25%.

At Pennsylvania State University, I led a team building customer engagement models that generated $500K in incremental revenue. I developed a predictive model with 87% accuracy by optimizing features from more than 1,600 variables.

My research and projects span network reliability, agricultural analytics, and AI systems. I developed an LSTM-based solution for predicting service disruptions and built a RAG pipeline for research papers using Python, LangChain, and ChromaDB.

Experience

Work history, roles, and key accomplishments

Current

Data Analyst

Current

LS Direct

Jun 2024 - Present (2 years 4 months)

Led enterprise chatbot linking ChatGPT to BigQuery via MCP, exposing governed tables, rules and regulations securely. Built predictive and segmentation models to analyze customer behavior across 50+ e-commerce clients, increasing ROI by 25%.

Python SQL BigQuery MCP ChatGpt predictive modeling Segmentation A B Testing Campaign Automation Reporting Pipelines

Data Science Research Associate

Pennsylvania State University

Aug 2023 - May 2024 (9 months)

Led 10-person team building customer engagement models that generated $500K incremental revenue. Built predictive model with 87% accuracy by optimizing features from 1,600+ variables, exceeding 10+ benchmarks.

Python predictive modeling Feature Engineering Featurewiz RFE SelectKBest Team Leadership

Data Science Intern

FarmwiseAI

Sep 2022 - Jan 2023 (4 months)

Built Python-based geospatial data pipelines enabling real-time agricultural analytics. Designed a scalable image processing framework reducing UAV dataset preparation time from hours to 5 minutes.

Python Geospatial Data Image Processing QA Processes Data Pipelines

MITACS Globalink Research Intern

Dalhousie University

May 2022 - Aug 2022 (3 months)

Improved network reliability predictions by 20% through PCA and feature engineering within a statistical analytical framework. Developed an LSTM-based predictive analytics solution with an 88% F1 score.

Python PCA Feature Engineering LSTM Time Series Analysis SMOTETomek Statistical Analysis

Education

Degrees, certifications, and relevant coursework

Pennsylvania State University

Master of Science, Data Analytics

2023 - 2025

Grade: 3.92

Activities and societies: Chancellor's Scholarship Awardee; Certification in Foundation of Data Science

Pursued a Master of Science in Data Analytics with a GPA of 3.92, and received the Chancellor's Scholarship Awardee and a Certification in Foundation of Data Science.