Richard Burton is a Staff Software Engineer in Cambridge with nine years’ experience building and deploying efficient ML solutions on Arm hardware, from Cortex-M microcontrollers to Ethos NPUs. He leads a team focused on NLP transformer integration for Android while continuing hands-on work in model training, quantization and pruning for edge deployment. His contributions to Arm’s ML demos and embedded evaluation kit translate bleeding-edge IP into partner-ready demos and practical guides. An active open-source contributor, he’s added TensorFlow Lite keyword-spotting examples and int16 quantization support to enable ML on resource-constrained devices. Trained as a mathematician, he combines rigorous numerical thinking with systems-level engineering to optimize models across software stacks and custom hardware.
9 years of coding experience
8 years of employment as a software developer
Master of Mathematics (MMath) Mathematics, Master of Mathematics (MMath) Mathematics at University of Leicester
Completed the CS231n 2017 spring assignments from Stanford university
Role in this project:
ML Engineer
Contributions:82 commits, 24 PRs, 73 pushes in 3 years 9 months
Contributions summary:Richard's contributions primarily focused on implementing and refining machine learning models within the cs231n-2017 repository. They implemented the SVM gradient calculation and vectorized the loss function for increased efficiency. The user also implemented the backpropagation for a 2-layer network and the dropout forward pass. Furthermore, they implemented the cross-entropy loss function for the softmax classifier and vectorized it.
Contributions:1 review, 8 commits, 10 PRs in 2 years 1 month
Contributions summary:Richard's commits focus on adding and improving examples related to keyword spotting on Cortex-M microcontrollers using TensorFlow Lite. These contributions involve modifications to data loading, model training, and model conversion processes, including adding support for new features like model conditioning and int16 quantization. This work aims to enable machine learning model deployment on resource-constrained embedded platforms.
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.