Chad EdwardsCo-founder and CEO · CuspAIFounder↗Both at CuspAI

Researcher · Academic · Founder · Executive
CuspAICo-founder and CTO
University of AmsterdamFull Professor
🇳🇱 DutchBorn 1968
Welling co-created variational autoencoders, which learn to generate new examples from data[1], and graph convolutional networks, which learn from relationships such as molecular bonds.[2] His work also helps models quantify uncertainty[3], and he co-founded CuspAI to apply AI to materials discovery.[4]
CuspAI2024-present
University of Amsterdam2012-presentWork and education history is primarily focused on AI-relevant roles and may not be comprehensive.
Variational autoencoders learn compact representations of data to generate new examples.
Original Kingma/Welling work introduces efficient variational learning of generative latent-variable models.
Original Kipf/Welling paper develops graph-convolution learning over graph structure and node features; molecular bonds are an illustrative graph relationship, not a claim about its original benchmark.
Original Welling/Teh paper introduces scalable Bayesian posterior sampling and explicitly discusses quantifying parameter uncertainty.
Current official company page identifies Welling as cofounder/CTO and describes materials-discovery purpose.
Based on public sources. If accuracy is important, verify details using the links or references.
For corrections, updates, or requests to remove information, contact us.
Reviewed August 8, 2026 · Checked for accuracy and new information.