Stephen Welch
Founder at WelchLabs
Winston-Salem, North Carolina, United States
Summary
🤩
Rockstar🎓
Top SchoolStephen Welch is a founder and engineering leader with 12 years of experience building production ML and computer vision systems and a parallel career creating widely viewed educational content. As founder of Welch Labs he authored a YouTube channel with 50M+ views and 550k+ subscribers (translated into 25+ languages), while in industry he led teams and shipped deep-learning inspection and perception products that delivered multimillion-dollar value to customers. He has held senior engineering and data science roles—VP Data Science at Mariner, VP Machine Learning at Autonomous Fusion, and Engineering Manager at Invisible AI—where he built training pipelines in PyTorch/fastai and lightweight architectures for real-time inference. His open-source work includes practical ML teaching code (Neural-Networks-Demystified) used to demystify forward/backprop and numerical gradient checking. With graduate training from UC Berkeley and an undergrad in electrical engineering, he blends academic rigor, product delivery, and a knack for translating complex ML concepts into accessible explanations.
12 years of coding experience
15 years of employment as a software developer
M.S. Environmental Engineering, M.S. Environmental Engineering at University of California, BerkeleyB.S. Electrical Engineering, B.S. Electrical Engineering at Georgia Institute of Technology
English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (6)
neural-network10
machine-learning10
gradient-descent10
python10
backpropagation10
numpy10
Programming languages (4)
Github contributions (5)
stephencwelch/Neural-Networks-Demystified
Jan 2015 - Mar 2020
Supporting code for short YouTube series Neural Networks Demystified.
Role in this project:
ML Engineer
ML EngineerContributions:36 commits, 9 PRs, 36 pushes in 5 years 3 months
Contributions summary:Stephen primarily contributed to the development of the neural network project by adding and updating code related to forward propagation, gradient descent, and backpropagation, which are all fundamental concepts in neural networks. They implemented methods and functions, including activation functions, and cost calculations, while fixing errors. The user also incorporated a numerical gradient-checking process to validate the implementation and ensure accurate results, and ultimately trained the network.
youtube-seriesdeep-learningneural-networksmachine-learningshort
stephencwelch/Neural-Networks-For-Audio-Processing
Feb 2015 - Aug 2017
Contributions:35 commits, 32 pushes, 1 branch in 2 years 6 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.