London Meetup Recording

Amplify Your Graph Neural Network Workflows

It was packed meetup in London with out partner Graphcore where we hosted an evening of talks from GNN industry leaders.  

Sign up here to receive the recordings.



"Amplify Your GNN Workflows"

  • Jure Leskovec, Professor at Stanford & Co-Founder at Kumo AI 
  • Matthias Fey, Founding Engineer at Kumo AI & Creator of PyTorch Geometric 

"Hardware for Graph Neural Networks"

  • Andrew Fitzgibbon, Technical Fellow at Graphcore

"PyTorch Geometric for IPUs"

  • Hatem Helal, Research Scientist at Graphcore

"Graph ML Research at Graphcore"

  • Dominic Masters, Technical Lead for GNNs at Graphcore





Jure Leskovec

Chief Scientist and Co-founder

Kumo AI

Jure is the Chief Scientist at and one of the leading AI and data mining professors at Stanford University. Jure was the former Chief Scientist at Pinterest.

Andrew Fitzgibbon

Andrew Fitzgibbon

Research Fellow



Andrew has been closely involved in the delivery of three groundbreaking AI systems over two decades. He was computer vision lead on the Emmy-award-winning 3D camera tracker “Boujou”; Kinect for Xbox 360 introduced large-scale synthetic training data to machine learning, and he was science lead on the team that shipped fully articulated hand tracking on HoloLens 2. His passion is bringing the power of mathematics to the crucible of real-world engineering. He has numerous research awards, including ten “best paper” prizes at leading conferences, and is a Fellow of the UK’s Royal Academy of Engineering. 

matthias frey

Matthias Fey

Founding Engineer

Kumo AI


Matthias is a founding engineer at where we work on making state-of-the-art Graph Neural Network solutions readily available to large-scale data warehouses. Previously, he obtained his PhD at the TU Dortmund University, Germany. Furthermore, he is the creator of the PyG (PyTorch Geometric) library, a widely used open-source library to easily write and train Graph Neural Networks for a wide range of applications related to structured data.

Hatem Helal
Dominic Masters

Hatem Helal

Research Scientist



Hatem Helal is a research scientist at Graphcore working on graph neural networks and quantum chemistry.  Previously he was responsible for scaling MATLAB algorithms to large distributed systems at MathWorks. Hatem holds a PhD in physics from the University of Cambridge and a BS in applied physics from Caltech.

Dominic Masters

Lead Research Scientist



Dominic Masters is a research scientist and the technical lead for graph neural networks at Graphcore. Throughout his six years at Graphcore he has led projects in areas such as memory efficient training methods, computer vision and graph machine learning. Most recently he led a team to first place in the Open Graph Benchmark Large Scale Challenge for molecular property prediction.



Graphcore is a UK-headquartered leader in next generation AI compute, underpinning the generative AI revolution. AI solutions based on Graphcore technology are already opening-up a world of possibilities - driving the creation of new AI-native product firms and transforming business practices at established companies. 
Graphcore’s made-for AI IPU technology is available in the cloud to deliver next-level compute efficiency, for commercial users, working at scale, who don’t want to choose between price and performance. Generative AI gives you the power to imagine something new. Graphcore AI cloud compute gives you the tools to build it. 



Kumo AI is a SaaS AI platform that allows anyone in an organization to tackle predictive problems in business. It enables all non-technical and technical users, regardless of ML experience, to traverse all the major steps of a best practice ML lifecycle. Graph learning can be operationalised in days not months, tackling a wide variety of predictive problems such as new customer acquisition, customer loyalty and retention, customer communications optimization, personalization, entity resolution and knowledge graph curation, abuse detection, financial crime detection. Kumo is the home of the open source project PyTorch Geometric.