AI and software engineer building privacy-preserving technology.

🔒 PySyft - Perform data science on data that remains in someone else's server

📡 Syft Client - Peer-to-peer data science and AI via channels your organization already trusts

🧪 Private ML - Machine Learning with SPDZ

🧰 Toolbox - A tool for downloading your data and using it via CLI's

🦑 pymemri - Python SDK for building plugins on the Memri personal data platform legacy

🗄️ pod - The open-source backend for Memri legacy

🔌 pyintegrators - Import, enrich, and act on personal data from Gmail, WhatsApp, and more legacy

📱 ios-application - The iOS front end of Memri legacy

🌐 browser-application - The browser front end of Memri legacy

Three Tools for Practical Differential Privacy - Practical tools for making differentially private ML actually work: sanity checks, adaptive clipping, and large-batch training. link

Syft 0.5: A Platform for Universally Deployable Structured Transparency - A framework combining privacy-enhancing technologies for universally deployable structured transparency. link

Training a CNN using SPDZ - How to train convolutional neural networks using secure multi-party computation with 6x reduced communication overhead. link

Secure Enclaves for AI Evaluation - How OpenMined, UK AISI, and Anthropic used secure enclaves to evaluate AI models while protecting proprietary assets. link