// independent deep-tech engineering

From signal
to system.

I work across the full signal chain: the sensors and electronics that read the physical world, the processing and models that make sense of it, and the control, automation and software that act on it. Research-grade depth, delivered as systems you can actually run, and validated to regulated standards when it counts.

  • physical ↔ digital
  • research → production
  • patent-granted methods
  • GMP-grade when required

// the through-line

One discipline, end to end.

Most projects break at the seams between hardware, software, data and compliance. I work the whole chain, so those seams disappear.

  1. 01

    Sense

    Sensors, RF, instrumentation. Capture the physical world as clean, trustworthy measurement.

  2. 02

    Process

    Signal processing, estimation, machine learning. Turn raw measurement into meaning.

  3. 03

    Act

    Automation, control and software. Build systems that respond and keep running.

  4. 04

    Assure

    Validation, data integrity, QA. Prove it works, and keep it inspection-ready.

// capabilities

Deep across an unusually wide stack.

Six domains I go properly deep in, and the rare ability to connect them in a single working system.

rf / iot

RF, Wireless Sensing & IoT

Sub-GHz sensor networks, device-free sensing and large-scale real-world deployments, with patented calibration for reliability across changing environments.

embedded

Embedded & Instrumentation

Firmware and electronics in C, from bare-metal prototype to production-grade embedded system, including the hardware-software boundary where most projects quietly stall.

dsp / ml

Signal Processing, Data & ML

Estimation models, measurement pipelines and applied machine learning that pull real signal out of noisy, messy data.

s7 / ot

Industrial Connectivity & OT Data

S7 communication with Siemens PLCs, industrial IoT, and moving trustworthy data off the plant floor into monitoring and validated systems, including inside Grade A to D regulated environments.

ai / agents

AI Agents & Automation

LLM-agent workflows for validation documentation, change-control preparation and audit-trail generation, built to run inside GxP data and access constraints, not around them.

csv / gmp

Validated & Regulated Engineering

Computerised system validation (GAMP 5), GMP, Annex 1 and data integrity (ALCOA+) for life-sciences manufacturing, when the work has to pass inspection.

// track record

Invents the method, then ships the system.

A PhD research background and a granted patent at one end; production systems that have served over a million users and validated plant-floor systems at the other. The rare engineer who is genuinely at home at both ends.

Research-grade depth
PhD
Users on production systems built & run
0
Regulated cleanroom grade
A–D
Data-integrity findings, by design
Zero

// engagement

Senior, hands-on, end to end.

One brain across the stack

Fewer handoffs and fewer seams. The person specifying the sensor is the person validating the system.

Risk-based, not paperwork-based

Full rigour where the risk actually is, and none of the ceremony where it isn't.

Research depth, production discipline

Novel methods where the problem demands them; boring, reliable engineering everywhere else.

Vendor-independent

The right architecture for your problem, not whatever a hardware catalogue is selling this year.

// where it helps

Built for hard problems in serious places.

Wherever measurement, control and compliance have to be right the first time.

Life sciences & pharma
Industry 4.0 & manufacturing
IoT & smart infrastructure
Telecom & connectivity
Medical devices & diagnostics
Deep-tech startups & R&D

// let's talk

Have a problem that crosses disciplines?

The ones that fall between hardware, software, data and compliance are exactly the ones I take on. Let's scope it while there's still room to do it properly.