Arpan Parikh
Senior Machine Learning Engineer at Prudential Financial
Atlanta, Georgia, United States
Summary
👤
Senior🎓
Top SchoolArpan Parikh is a Senior Machine Learning Engineer based in Atlanta with nine years of experience building production ML and software systems. He has designed and deployed generative AI solutions and ML endpoints at Prudential using AWS SageMaker, and previously developed high-frequency monitoring and CI improvements across fintech stacks using Golang, Java, C#, Jenkins, Grafana and Prometheus. An active open-source contributor, Arpan enhanced PyTorch-Ignite with robust NLP and GAN metrics (ROUGE, IS, FID) implemented for distributed and online evaluation, reflecting a strong emphasis on reproducible model evaluation. His academic work spans ML for brain–computer interfaces, human action recognition, and an optimized attention-based text summarization thesis that halved training time while preserving ROUGE performance. Comfortable from low-level C/C++ to Python and JavaScript, he blends research rigor from Georgia Tech and BITS Pilani with pragmatic production engineering. Notably, he has a knack for turning evaluation math into stable, online metrics that scale in distributed training environments.
9 years of coding experience
2 years of employment as a software developer
BITS Pilani, Birla Institute of Technology and ScienceMaster of Science - MS Machine Learning in Computer Science, Master of Science - MS Machine Learning in Computer Science at Georgia Institute of Technology
English, Hindi, Gujarati, French
Github Skills (10)
pytorch10
deeplearning-ai10
machine-learning10
deep-learning10
python10
metric10
testing10
nlp9
computer-vision8
tensorflow4
Programming languages (4)
Github contributions (5)
pytorch/ignite
Mar 2021 - Aug 2021
High-level library to help with training and evaluating neural networks in PyTorch flexibly and transparently.
Role in this project:
ML Engineer
ML EngineerContributions:65 reviews, 9 commits, 26 PRs in 5 months
Contributions summary:Arpan's contributions centered on enhancing the `pytorch/ignite` library with new metrics, specifically Rouge and Inception Score, essential for evaluating machine learning models. They implemented and refined these metrics, including handling multi-reference scenarios and addressing potential numerical issues. Furthermore, the user focused on integrating and testing these metrics within the PyTorch framework, updating documentation and ensuring compatibility across different PyTorch versions and distributed environments. This work directly supports the project's goal of providing tools for training and evaluating neural networks.
pytorchpythondeep-learninghigh-levelneural-networks
gucifer/ignite
Apr 2021 - Aug 2021
High-level library to help with training and evaluating neural networks in PyTorch flexibly and transparently.
Contributions:174 pushes, 27 branches in 4 months
pytorchdeep-learninghigh-levelneural-networksmachine-learning
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.