Sumit Pai
Senior Consultant (Sr. Tech. Mgr.) - ML GenAI Knowledge Graphs at Visa
Bengaluru, Karnataka, India
Summary
🤩
Rockstar🎓
Top SchoolSumit Pai is a Senior Consultant and ML engineering leader based in Bengaluru with a strong 14+ year foundation in deep learning and 8 years of focused experience building production-grade GenAI, retrieval-augmented, and knowledge-graph solutions for enterprises. He has led R&D and engineering efforts at Accenture Labs, Deloitte, and now Visa, shipping scalable pipelines for metadata extraction, RAG workflows, document-search agents, and enterprise knowledge management. An active contributor to the AmpliGraph open-source library, he optimized GPU negative sampling and added regularization techniques to improve knowledge-graph embeddings—a practical improvement that reflects his blend of research and systems-level optimization. Sumit’s work spans adapting open-source LLMs with parameter-efficient tuning, publishing in top ML venues, and applying graph embeddings for anomaly detection across regulated domains, demonstrating both academic rigor and applied impact.
8 years of coding experience
7 years of employment as a software developer
BE, Computer Science and Engineering, BE, Computer Science and Engineering at NMAM Institute of TechnologyMaster of Science (M.Sc.), Neural Information Processing, Master of Science (M.Sc.), Neural Information Processing at University of Tuebingen
Stackoverflow
37reputation
27kreached
1answer
0questions
Github Skills (9)
machine-learning10
knowledge-graph10
tensorflow10
gpu10
python10
model-optimization10
regularization10
eclipse6
jar6
Programming languages (2)
Github contributions (5)
Accenture/AmpliGraph
Dec 2018 - Oct 2022
Python library for Representation Learning on Knowledge Graphs https://docs.ampligraph.org
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:3 reviews, 297 commits, 35 PRs in 3 years 11 months
Contributions summary:Sumit significantly enhanced the performance and functionality of the `ampligraph/ampligraph` repository. They refactored core code related to negative generation, migrating it to the GPU to improve performance, involving changes to the `evaluation` and `models` modules. Additionally, they added both L1 and L2 regularization techniques to the model for enhanced accuracy and robustness. These changes indicate an effort to optimize and refine the machine learning models.
python-libraryrepresentationpythonknowledgerepresentation-learning
sumitpai/test_repo
Feb 2019 - Feb 2019
Contributions:8 pushes, 1 branch in 1 day
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.