Adrian Lundell
Lund, Skåne County, Sweden
Summary
🤩
RockstarAdrian Lundell is an engineer with 8 years of experience specializing in embedded ML deployment and optimization, currently working at Arm in Lund. He focuses on integrating and tuning Executorch and CMSIS-NN for TensorFlow Lite Micro/LiteRT to bring efficient inference to resource-constrained devices. His open-source contributions to the widely used tflite-micro project include stabilizing Cortex-M builds, adding int16 CMSIS-NN LSTM support, and fixing critical buffer and dependency issues—work that directly improves edge ML reliability. Adrian’s background spans scientific programming at NASA Goddard, high-performance Bayesian parameter tuning for PRACE, and industrial tooling at ABB, giving him a strong mix of research, HPC and product engineering experience. With a master’s in Complex Adaptive Systems and early training in engineering mathematics and physics, he combines theoretical rigor with practical low-level systems know-how. Colleagues describe him as a pragmatic problem-solver who quietly improves performance and maintainability in complex embedded ML stacks.
8 years of coding experience
Github Skills (15)
cmsis10
embedded10
c-language10
cprogramming-language10
n10
sys10
architecture9
architectures9
machine-learning9
tensorflow-lite9
arm9
git7
bash7
githubaction-workflow5
github-ci5
Programming languages (4)
Github contributions (5)
tensorflow/tflite-micro
Jun 2023 - Nov 2024
Infrastructure to enable deployment of ML models to low-power resource-constrained embedded targets (including microcontrollers and digital signal processors).
Role in this project:
Embedded Systems Engineer / IoT Developer
Embedded Systems Engineer / IoT DeveloperContributions:1 review, 15 PRs, 7 comments in 1 year 5 months
Contributions summary:Adrian primarily focused on enhancing the functionality and stability of the TensorFlow Lite for Microcontrollers project. Their contributions include updating external library dependencies like CMSIS-NN, fixing build errors related to Cortex-M architecture, and integrating support for new features like int16_t data types within the CMSIS-NN LSTM kernel. Additionally, the user addressed critical issues, such as fixing a read of a non-initialized buffer and updating download links for the Ethos-U platform, improving the overall performance and compatibility of the project. They also updated CMSIS-NN calls and optimized the project's code base by removing compiler options.
signalml-modelslow-powerprocessorsdeployment
AdrianLundell/executorch
Oct 2024 - Mar 2025
On-device AI across mobile, embedded and edge for PyTorch
Contributions:54 pushes, 35 branches in 5 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.