Gustav Larsson
Principal Engineer Manager at Qualcomm
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolGustav Larsson is a Principal Engineering Manager based in San Francisco with 14 years of experience building and shipping machine learning systems from research to edge deployment. He combines deep research credentials (PhD-level computer science from University of Chicago) with product-focused engineering, having led ML teams at Apple and a founding engineering role at Tetra AI that was acquired by Qualcomm. At Qualcomm he now drives edge AI deployment work that scales models to mobile devices, drawing on hands-on contributions to prominent open-source projects like PyTorch, CoreMLTools, and Turi Create where he improved model conversion, quantized ONNX export, and image-classifier reliability. Known for pragmatic engineering rigor, he focuses on performance, reproducibility, and reducing deployment friction—often fixing subtle type/shape issues and test flakiness that block production. Colleagues rely on him to bridge research, tooling, and product needs to get models into constrained real-world environments.
14 years of coding experience
19 years of employment as a software developer
Master of Science (M.S.) Engineering Physics, Master of Science (M.S.) Engineering Physics at KTH Royal Institute of TechnologyDoctor of Philosophy (Ph.D.) Computer Science, Doctor of Philosophy (Ph.D.) Computer Science at University of Chicago
Swedish, English, Japanese
Stackoverflow
8,329reputation
1.2mreached
103answers
10questions
php
top-5%
xna
top-5%
python
top-5%
Github Skills (34)
pytorch10
python10
image-classification10
scikit10
machine-learning10
onnx10
coreml10
deep-learning10
tensorflow10
scikit-learn10
quantization10
model-conversion10
convert9
converter9
testing9
Programming languages (11)
Github contributions (5)
apple/turicreate
Dec 2017 - Nov 2018
Turi Create simplifies the development of custom machine learning models.
Role in this project:
ML Engineer
ML EngineerContributions:55 commits, 82 PRs, 31 pushes in 11 months
Contributions summary:Gustav primarily contributed to the `turicreate` repository by addressing issues in the evaluation tests and improving the functionality of the image classifier. Their work involved fixing bugs related to scoring, particularly with the fbeta score and single image predictions. They also introduced a seed argument in the image classifier for reproducible results, and refactored code for compatibility.
pythondeep-learningmachine-learningcoremlmulticlass-classification
apple/coremltools
Oct 2017 - Apr 2018
Core ML tools contain supporting tools for Core ML model conversion, editing, and validation.
Role in this project:ML Engineer
ML EngineerContributions:34 commits, 30 PRs, 14 pushes in 5 months
Contributions summary:Gustav primarily contributed to the `coremltools` repository by addressing bugs and enhancing the tools related to machine learning model conversion and prediction. Their work included fixing type checking issues, improving Python 3 support, and marking slow tests. They also focused on addressing issues related to coreml model prediction on macOS.
pytorchmodel-conversionvalidationdeep-learningmachine-learning
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.