Jeff Ward
City Of Oak Harbor Planning Commission Position 1 at City of Oak Harbor, WA
United States
Summary
👤
Senior🎓
Top SchoolJeff Ward is a machine learning-focused engineer with eight years of industry experience and a diverse background that spans military aviation and advanced academic training (MSc Informatics, University of Edinburgh). He has built production AI/NLP systems at Deepgram, DeepScribe, and now leads as Principal Engineer at Humata Health, bringing practical GPU and PyTorch backend optimizations to open-source projects like deepgram/kur. Comfortable moving between low-level performance tuning and product-facing ML features, Jeff pairs hands-on engineering with systems thinking to deliver efficient, deployable models. Outside of tech he serves on the City of Oak Harbor Planning Commission, channeling his problem-solving skills into civic engagement and advocacy—an unusual blend of operational discipline and creative "crazy ideas" implementation.
8 years of coding experience
10 years of employment as a software developer
Master’s Degree, Informatics, With Merit, Master’s Degree, Informatics, With Merit at The University of EdinburghBachelor of Science - BS, Computer Science, Bachelor of Science - BS, Computer Science at San Francisco State University
Github Skills (10)
pytorch10
backend10
gpu10
back-end-development10
python10
optimisation10
optimization10
machine-learning9
deep-learning9
neural-network8
Programming languages (1)
Github contributions (5)
deepgram/kur
Sep 2017 - Nov 2020
Descriptive Deep Learning
Role in this project:
Back-end Developer
Back-end DeveloperContributions:5 commits, 2 PRs, 3 pushes in 3 years 2 months
Contributions summary:Jeff primarily focused on enhancing the backend functionality of the project, particularly in relation to GPU utilization and performance. Their work included adding configuration options for GPU busy percentages and integrating missing merge functions for the PyTorch backend. They also made adjustments to the codebase, such as allowing multiple outputs from a layer in the PyTorch backend and patching for Python 3.7 compatibility. These modifications suggest an effort to improve the system's efficiency and compatibility.
caffe2image-recognitiondeep-learningspeech-to-textneural-networks
deepgram/deepgram-brain-python
Oct 2017 - Mar 2018
Python API wrapper for the Deepgram API. Oh yes, it's cool. You should get it.
Contributions:1 release, 22 commits, 1 PR in 5 months
apipythonapi-wrapperyespython3
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.