Andrew Trask
Founder & Leader, OpenMined | Senior Research Scientist, Google DeepMind | Oxford PhD
Founder and executive leader of OpenMined, Senior Research Scientist at Google DeepMind, and PhD in Computer Science from the University of Oxford. Pioneer of privacy-preserving machine learning, federated learning, differential privacy, and encrypted computation. Author of the acclaimed book 'Grokking Deep Learning', training tens of thousands of software developers in neural network fundamentals from scratch.
Areas of focus
Professional niches
Proof of Work
Grokking Deep Learning
Authored the seminal deep learning curriculum teaching backpropagation, matrix calculus, convolution, and recurrent architectures using only standard Python and NumPy without frameworks.
Pedagogical pure NumPy implementations lack GPU-accelerated cuDNN kernels for enterprise-scale foundation model pretraining.
Context: Pedagogical algorithmic synthesis, foundational vector mathematics.
View missionPySyft: Library for Secure & Private Machine Learning
Engineered an open-source decentralized privacy framework extending PyTorch and NumPy with Federated Learning, Differential Privacy, and Multi-Party Computation (SMPC) across untrusted nodes.
Encrypted tensor computation induces substantial network latency and memory overhead compared to plain-text distributed SGD.
Context: Python, PyTorch, cryptographic SMPC protocols.
View mission