Ryan Chesler
Senior Applied Research Scientist at NVIDIA
San Diego, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRyan Chesler is a Senior Applied Research Scientist with nine years of experience building multimodal ML systems and shipping production-ready solutions across startups and enterprise teams. A self-taught data scientist and two-time Kaggle Grandmaster, he pairs competitive modeling expertise with practical engineering—most recently contributing vision and document-understanding integrations to the widely used h2ogpt project. He has held senior and principal roles at H2O.ai, Deep Sentinel, and NVIDIA, focusing on multimodal retrieval, computer vision, and private local model deployments. Based in San Diego, he organizes the local machine learning meetup, translating community insights into applied research and product features. Comfortable moving from research prototypes to deployable systems, he brings a pragmatic, privacy-minded approach to multimodal AI that emphasizes real-world document and image understanding.
9 years of coding experience
10 years of employment as a software developer
Bachelor's degree Business Entrepreneurship, Bachelor's degree Business Entrepreneurship at Arizona State University
Stackoverflow
45reputation
3kreached
0answers
3questions
Github Skills (16)
computer-vision10
machine-learning10
python10
image-processing10
llm9
generative-ai9
pdf-processing9
ai7
gpt7
htaccess6
php6
tensorflow6
embedding6
magento6
mixture5
Programming languages (2)
Github contributions (5)
h2oai/h2ogpt
Apr 2023 - Apr 2024
Private chat with local GPT with document, images, video, etc. 100% private, Apache 2.0. Supports oLLaMa, Mixtral, llama.cpp, and more. Demo: https://gpt.h2o.ai/ https://gpt-docs.h2o.ai/
Role in this project:
ML Engineer
ML EngineerContributions:3 reviews, 8 PRs, 22 pushes in 11 months
Contributions summary:Ryan's commits primarily focus on integrating and enhancing image processing capabilities within the h2ogpt project. They added support for H2O DocTR, a document text recognition tool, and integrated it for PDF processing. The user also incorporated and refined functionalities for Blip and Pix2Struct image analysis and captioning, demonstrating expertise in computer vision and document understanding within the context of the project's goal of private, local GPT interaction with multimedia. These changes suggest an effort to enhance the system's ability to handle and interpret visual content, reflecting a focus on enriching the types of documents the system can process.
aichatgptembeddingsgenerativegpt
ryanchesler/NN-Plot
Nov 2017 - Nov 2017
Contributions:3 commits, 2 pushes, 1 branch in 10 days
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.