James Thewlis
Co-Founder at Unitary
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolJames Thewlis is a co-founder and machine learning engineer with 13 years of experience building and deploying computer vision and deep learning systems from research to product. Based in London, he bridges academic rigor (DPhil work with Oxford’s Visual Geometry Group and publications with Facebook AI Research) and startup execution at Unitary and through Entrepreneur First. His hands-on contributions to the widely used Caffe framework include GPU kernel optimization, multi-GPU setup, and data-handling fixes, reflecting deep low-level understanding of ML infrastructure. He has applied that expertise commercially as a consultant and specialist for Mirriad, turning state-of-the-art object and scene recognition into production solutions. Notably, his background combines GPU/CUDA implementation experience (face detection) with leadership in founding an ML-focused company, making him equally fluent in research, code, and product delivery.
13 years of coding experience
2 years of employment as a software developer
DPhil AIMS CDT and Visual Geometry Group Supervisor: Andrea Vedaldi, DPhil AIMS CDT and Visual Geometry Group Supervisor: Andrea Vedaldi at University of OxfordMEng Computing, MEng Computing at Imperial College London
Spanish, French, Romanian, English
Github Skills (17)
caffe10
c-language10
gpu-programming10
machine-learning10
machine-vision10
deeplearning-ai10
vision-api10
deep-learning10
cuda10
computer-vision10
cprogramming-language10
testing9
convolutional-neural-networks9
neural-network9
object-detection8
Programming languages (9)
Github contributions (5)
BVLC/caffe
Mar 2014 - Jun 2014
Caffe: a fast open framework for deep learning.
Role in this project:
ML Engineer
ML EngineerContributions:26 commits, 1 comment in 3 months
Contributions summary:James primarily contributed to the development and testing of the core deep learning framework. Their commits focused on fixing bugs related to data handling within the network, improving the functionality of the tools used for network inspection, and optimizing GPU kernel operations. They also worked on the proper function calling, and the setup for using multiple GPUs. The user demonstrated an understanding of the underlying Caffe library by implementing and testing a new im2col kernel.
caffedeep-learningmachine-learningvision
jamt9000/matconvnet-rcnn
Apr 2015 - Jul 2016
Contributions:33 commits, 27 pushes, 5 branches in 1 year 3 months
r-cnncnnmatconvnet
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.