Diego Antognini
Senior Research Engineer at Lucerne University of Applied Sciences and Arts
Zurich, Zurich, Switzerland
Summary
🤩
Rockstar🎓
Top SchoolDiego Antognini is a Senior Research Engineer and tech lead at Google DeepMind with 12 years of experience building and deploying AI systems, currently contributing to Gemini. He combines deep academic training (PhD, EPFL) and hands-on industry research from IBM to DeepMind, with a track record of publications, patents, and low-latency model engineering for production use. Diego teaches and supervises MSc students in NLP and LLMs, having designed courses and guided 190+ students and 20+ theses across applied domains like medicine, law, and finance. His contributions span graph neural networks (PyTorch GAT implementation) to scalable pipelines that turn conversations into SQL and tiny, high-throughput term encoders deployed in real products. Colleagues know him for bridging rigorous research with practical engineering—optimizing models for CPU latency and real-time inference while mentoring the next generation of ML practitioners. Based in Zurich, he blends academic rigor, product-focused invention (multiple patents), and open-source contributions to advance applied AI.
12 years of coding experience
3 years of employment as a software developer
CFC computer scientist + MPT Computer Science, CFC computer scientist + MPT Computer Science at Centre Professionnel du Littoral NeuchâteloisDoctor of Philosophy (Ph.D.) Computer Science, Doctor of Philosophy (Ph.D.) Computer Science at EPFLBachelor of Science HES-SO Computer Science, Bachelor of Science HES-SO Computer Science at Haute Ecole Arc - Ingénierie
French, English, Spanish, German, Italian
Github Skills (10)
neural-network10
attention-mechanism10
pytorch10
machine-learning10
deep-learning10
python10
data-structure7
data-structures7
algorithm7
algorithms7
Programming languages (6)
Github contributions (5)
Diego999/pyGAT
Mar 2018 - Aug 2021
Pytorch implementation of the Graph Attention Network model by Veličković et. al (2017, https://arxiv.org/abs/1710.10903)
Role in this project:
ML Engineer
ML EngineerContributions:27 commits, 8 PRs, 21 pushes in 3 years 6 months
Contributions summary:Diego primarily contributed to the development and training of a Graph Attention Network model. Their work involved implementing the GAT architecture using PyTorch, including the GraphAttentionLayer and GAT modules. They also focused on data loading, model training, and evaluation, incorporating techniques like dropout, weight decay, and early stopping. The user's changes included refactoring code, adding Xavier initialization, and correcting the normalization of the adjacency matrix.
pytorchthe-graphpythonarxivabs
Diego999/SelfSent
May 2017 - Jul 2017
Contributions:44 commits, 12 pushes, 1 comment in 1 month
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.