Edge AI · Embedded ML

I build intelligence that runs
on the device, not the cloud.

Embedded-ML engineer specialising in on-device inference: voice agents, sensor fusion, and real-time control that work with no connection and no data centre.

Try the live van demo → See the work

Open to edge-AI and embedded-ML roles, and broader ML or systems work where on-device thinking helps. Remote preferred, happy to work hybrid or relocate for the right team. Available now.

What I do

On-device by default

VOICE

Local voice agents

Speech in, action out. ASR, an LLM agent loop, and TTS running on a single edge module with no round-trip to the cloud.

SENSING

Telemetry & fusion

Reading real hardware (power systems, GPS, cameras) and turning noisy signals into something a person can act on.

SYSTEMS

Whole systems

From the driver layer to the web UI: testable, lightweight builds that survive flaky power and patchy connectivity.

About

I live in the system I built

After a PhD in embedded machine learning, I went looking for the hardest place to make AI actually useful: off-grid, no cloud, flaky power, one small board doing everything. So I built it into a campervan, and now I live in it full-time. The same instinct runs through my work. I put intelligence where the problem is, on the device, in the field, not in a data centre.

Featured

The Smart Van

A bespoke edge-AI control system on a Jetson Orin Nano. Local voice control, live Victron power telemetry, GPS and location, 4G with data-usage tracking, a camera, lights, and SMS, all running on-device with no cloud in the loop. The demo below is the real dashboard running on simulated data.

Open the demo Read the case study