Himanshu Chaudhary

Data Scientist

Madison, Wisconsin, United States3 yrs 9 mos experience

AI ML PractitionerAI Enabled

Key Highlights

Expert in machine learning engineering and AI research.
Developed models improving fraud detection and customer engagement.
Strong academic background with MS in Computer Science.

Stackforce AI infers this person is a Data Scientist specializing in Fintech and AI-driven solutions.

Contact

Skills

Core Skills

Machine Learning

Other Skills

Deep LearningData ScienceScikit-LearnXGBoostModelingSQLApache SparkApache ImpalaPySparkBERT (Language Model)MistralModel TrainingLarge Language Models (LLM)Generative AIPyTorch

About

Data Scientist at PayPal (focused on machine learning engineering), with an MS in Computer Science from UW–Madison. I bring 3+ years of industry experience at Mastercard and PayPal, working on fraud detection, recommendation systems, and applied generative AI — including production-grade BERT-based and LLM-powered models. My research has focused on integrating large language models (LLMs) with graph neural networks (GNNs) to enhance reasoning over structured knowledge. I’m passionate about building scalable ML systems and advancing applied AI research.

Experience

Paypal

2 roles

Data Scientist

Jun 2025 – Present · 9 mos · New York City Metropolitan Area · Hybrid

Machine Learning Intern

May 2024 – Aug 2024 · 3 mos · New York City Metropolitan Area · Hybrid

- Utilized unsupervised clustering with Mistral 7b to uncover underlying sentiments in merchant conversations with loan salespersons. Developed a RoBERTa-based model achieving a 76% F1 score for the automated classification of multi-turn conversations. Discovered previously unrecognized reasons for walkaways, which are now being used to refine sales strategy and improve customer engagement.

Deep LearningMachine Learning

University of wisconsin-madison

Graduate Student Researcher

Jan 2024 – Jun 2025 · 1 yr 5 mos · On-site

Working with Dr. Junjie Hu (https://junjiehu.github.io/) on a novel GAT-based architecture for joint LLM-graph modeling to improve Graph Question Answering by enabling better reasoning over structured graph knowledge and natural language.

Deep LearningMachine Learning

Mastercard

2 roles

Data Scientist 2

Aug 2022 – Aug 2023 · 1 yr

- Worked on creating new set of graph based features for the existing fraud detection model. Used personalized pagerank algorithm on Merchant-Users-Merchant connected graph to generate the features. The new set of features improved the model performance by atleast 5%.

Machine LearningData Science

Data Scientist 1

Aug 2020 – Aug 2022 · 2 yrs

Built state-of-the-art hierarchical Bert-based model to enable transfer learning for transactional data. Employed masked language modelling and random token detection strategy for pretraining the model. Fine tuned the pretrained model for one of the existing use case to beat the baseline model performance.
Built recommendation models to identify and assist customers to amplify their revenue with help of targeted campaigns, leading to an estimated increase by $30 million worth of transactions monthly.

Samsung r&d institute bangalore

Summer Intern

May 2019 – Jul 2019 · 2 mos · Bengaluru, Karnataka

Worked under CMC server team. The purpose of the CMC server is to integrate multiple secondary devices (without sim) with one primary device(with sim) to synchronise the Call and Message log on all the devices. Added new tables to the server’s database and added a new API to the server.