Kshitij Fadnis
Research Engineer at IBM
City of White Plains, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolKshitij Fadnis is a research engineer at IBM Research with nine years of experience building conversational AI, NLP and ML systems, grounded in a Master’s in Computer Science from Ohio State. He specializes in deep reinforcement learning, domain-independent symbolic reasoning, and production-grade NLP libraries, and has contributed backend enhancements to IBM’s popular pytorch-seq2seq framework around checkpointing, training and evaluation. At IBM he has moved between core Watson algorithm work, search and dialog services, and analytics microservices, pairing research rigor with shipping enterprise solutions and mentoring teams. Based in White Plains, NY, he blends interests in cognitive science, computer vision and robotics to bridge symbolic and neural methods—often improving reliability and tooling behind conversational systems rather than just model architectures.
9 years of coding experience
3 years of employment as a software developer
Bachelor of Engineering, Electronic Instrumentation, Bachelor of Engineering, Electronic Instrumentation at University of MumbaiMaster’s Degree, Computer Science and Engineering, Master’s Degree, Computer Science and Engineering at The Ohio State University
English, Hindi, Marathi
Github Skills (5)
pytorch10
machine-learning10
python10
seq2seq10
testing9
Programming languages (5)
Github contributions (5)
IBM/pytorch-seq2seq
Jul 2017 - Aug 2017
An open source framework for seq2seq models in PyTorch.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:21 commits, 14 PRs, 28 pushes in 1 month
Contributions summary:Kshitij's contributions primarily revolve around modifying and maintaining core components related to checkpointing, training, and evaluation within the seq2seq framework. These changes include modifications to the checkpointing mechanism, such as saving and loading models and optimizer states. Furthermore, the user made adjustments to the supervised training process, likely to enhance efficiency or incorporate new features related to data handling and batch processing. The changes also include testing and debugging of the codebase.
deep-learningpytorchvisual-recognitionseq2seq
IBM/InspectorRAGet
Apr 2024 - Mar 2025
The repository contains generative AI analytics platform application code.
Contributions:6 reviews, 16 PRs, 46 pushes in 11 months
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.