Hello, I'm Carrow.

I'm a Full-Stack AI Engineer and Data Architect finishing a PhD in Urban Digital Twins at Newcastle University's Geospatial Systems CDT. My research designs hybrid Graph Attention Networks and Transformer encoders for IoT traffic sensor data — work focused on making spatial deep learning resilient to the messy, irregular reality of production sensor networks.

In parallel, I'm a Turing Enrichment Student at The Alan Turing Institute, developing “TransformerNP” — a new Neural Process paradigm for the DeepSensor framework that accelerates training and inference for environmental forecasting. The work runs on HPC clusters via SLURM with centralised MLOps in Weights & Biases for the wider research team.

Alongside the research, I build production systems. As Technical Lead on the NOVA Digital Twin for the Dept. for Business and Trade, I architected a scenario-based Multi-Criteria Evaluation model and interactive digital-twin features for renewable siting. As solo architect of an Open Banking financial dashboard, I shipped a TypeScript monorepo with NestJS, PostgreSQL/Prisma, and WebAuthn Passkey 2FA — ingesting a decade of live transaction data under a restricted Plaid Production Licence.

The throughline is a builder mentality and a tenacious interest in how complex systems behave — across finance, urban systems, and geospatial intelligence. I write requirements first, design decoupled architectures where each component can be upgraded independently, and lean on infrastructure-as-code, CI/CD, and automated testing to keep systems trustworthy as they grow.

I'm based in Newcastle upon Tyne, relocating to London, and available for hybrid or on-site roles. If you're building data products for complex domains, I would love to talk.