I’m an AI specialist with a background in research science, applied AI engineering, and software engineering. I’m currently Founder & CEO of Spatial Intelligence - a deeptech AI startup building tools for machines that understand the structure and content of 3D environments.
My work spans the gap between fundamental research and production systems - from designing self-supervised world models and reasoning architectures, to shipping scalable AI-powered products on cloud infrastructure.
I care about outcomes over output and impact over influence.
I enjoy providing (and receiving!) mentorship and thrive in tightly-knit, high-agency interdisciplinary teams.
I think the best researchers and engineers are creatives who are genuinely fascinated by the technology they build, and humble enough to keep learning from everyone around them.
I strive to model myself around these values, and to share them with others.
Work
My research centres on spatial intelligence - specifically, how machines can learn structured, geometrically grounded representations of the world without explicit supervision. Active areas include self-supervised representation learning (JEPAs), object-centric world modelling, and hierarchical latent-space reasoning (Latent Cycle Transformers). Applied work includes spatially-grounded AI agents and 3D scene generation from text descriptions.
I build across the full stack - Python APIs, cloud infrastructure (Azure, AWS), ML systems, and TypeScript frontends - with a focus on reliable, scalable AI products. Current engineering work includes an async, GPU-bound, multi-stage 3D scene generation pipeline using LLM orchestration, vector search, and event-driven autoscaling on Azure Container Apps.
Education
PhD (Discontinued)
PhD in Artificial Intelligence, Imperial College London, 2022-2022 - Discontinued
MEng
MEng in Mechanical Engineering, University of Bristol, 2017-2021 - First Class Honours
Always tinkering - everything is figure-out-able…