Gustaf Ahdritz is a PhD candidate and software engineer with seven years of experience building production-ready ML systems and MLOps pipelines from Cambridge, MA. He brings deep hands-on expertise in GPU-efficient PyTorch engineering, notably contributing to openfold—a trainable, memory-efficient AlphaFold2 reproduction—where he implemented core modules like Invariant Point Attention and memory-aware inference. Gustaf combines research rigor with practical backend and deployment skills, bridging academic models and scalable tooling. Educated at Columbia, he operates comfortably at the intersection of computational biology and systems engineering. Colleagues value his ability to translate complex model components into maintainable, high-performance code and reproducible training workflows.
Trainable, memory-efficient, and GPU-friendly PyTorch reproduction of AlphaFold 2
Role in this project:
Back-end Developer & MLOps Engineer
Contributions:6 releases, 1 review, 674 commits in 1 year 3 months
Contributions summary:Gustaf's contributions appear to focus on developing a trainable, memory-efficient, and GPU-friendly PyTorch implementation of AlphaFold 2. Their commits involve adding and modifying code related to the core structure module, including implementing the Invariant Point Attention (IPA) and AngleResnet components. Further contributions encompass defining and integrating loss functions for the model, as well as setting up a memory-efficient inference pipeline.
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.