Emanuele Rossi
Postdoctoral Researcher at Amboss Technologies
Barcelona, Catalonia, Spain
Summary
🤩
Rockstar🎓
Top SchoolEmanuele Rossi is a graph deep learning researcher and postdoctoral researcher with 11 years of experience bridging cutting-edge research and production ML, currently based in Barcelona and affiliated with Sapienza Università di Roma. He completed a PhD at Imperial College and has industrial research experience at Twitter (and Fabula AI), where he helped build Temporal Graph Networks (TGN) and contributed significant refactors and performance improvements to the PyTorch Geometric ecosystem. His work spans temporal/edge-aware GNN architectures, fast inference engineering, and robust training pipelines—skills he has applied both in research internships (UCLA, Imperial) and industry roles. As a co-founder of a mentorship initiative and an advisor to ML startups, he pairs technical depth with mentorship and product sensibility. A less obvious strength is his knack for making research code reusable and production-ready, evidenced by TGN integrations and API-focused refactors in a major open-source library.
11 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Imperial College LondonMaster of Philosophy - MPhil Computer Science, Master of Philosophy - MPhil Computer Science at University of CambridgeLiceo Scientifico Rambaldi Valeriani, Imola
English, Italian
Github Skills (10)
geometric-deep-learning10
pytorch10
machine-learning10
deep-learning10
graph-neural-network10
python10
gnn10
modeling9
trainings9
graph-convolutional-networks9
Programming languages (8)
Github contributions (5)
twitter-research/tgn
Aug 2020 - Dec 2020
TGN: Temporal Graph Networks
Role in this project:
ML Engineer
ML EngineerContributions:3 reviews, 38 commits, 5 PRs in 4 months
Contributions summary:Emanuele contributed significantly to the implementation and refinement of the Temporal Graph Networks (TGN) model, as evidenced by the addition of TGN code and associated functionalities. Their work involved integrating random node feature randomization, adding non-linearity to the graph sum embedding module, and incorporating support for DyRep. The user also refactored and fixed bugs in the training scripts, focusing on node classification tasks.
graph-networksgraph-convolutional-networkstemporalgraph-representation-learninggraph-neural-networks
pyg-team/pytorch_geometric
Nov 2020 - Dec 2020
Graph Neural Network Library for PyTorch
Role in this project:ML Engineer
ML EngineerContributions:4 reviews, 16 commits, 1 PR in 20 days
Contributions summary:Emanuele primarily contributed to the development of a Temporal Graph Network (TGN) model within the PyTorch Geometric library. Their work involved refactoring the TGN model for improved API design, reusability across training and testing, and faster inference times. Key contributions include implementing a Graph Attention Embedding module, refactoring time encoding, and integrating edge features, demonstrating a focus on enhancing the model's performance and architecture.
pytorchgraph-convolutional-networksgeometric-deep-learningdeep-learningneural-graph
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.