William Jones

Decision-making depends on trust, but how often can you trace a reported figure back to its source? We're solving that with an open-source platform to repeatedly analyse any dataset against any geometry cloud-natively, with an auditable record of every step and full provenance fo…

During 2025 we ported ExecuTorch, the extension to PyTorch for embedded systems, to a bare metal multi-core RISC-V microcontroller based on the CORE-V CV32E40Pv2 processor. In this talk we'll explain the steps we had to take to achieve this.

Accelerator and Quantization Support on ExecuTorch - Chen Lai, Meta Agents in Space - George Williams, Armijo Innovations Enabling End-to-end Embedded ML Applications With ExecuTorch for the Arm Ethos-U NPU - George Gekov, Arm ExecuTorch for ML Inference on Edge Devices Running Zephyr RTOS - Bujji...

These days it is easy to design your own AI hardware accelerators using cheap FPGA boards and open source tooling. It's even possible to turn these into ASICs using open source technology.

This talk provide a gentle introduction to AI and Machine Learning, and the open source ecosystem around surrounding it. We start from the beginning, with no pre-existing knowledge is assumed.

Dynamic Causal Modeling is a uncertainty aware, explainable AI technique that uses physics inspired "dynamical systems" to explore time series data. In this talk we discuss the theory and practice of Dynamic Causal Modeling, and the work we've done to take this code outside of it's research rooted MATLAB implementation into a robust, general piece of software under a FOSS license.