pyPLECS
Python tooling that turns one-off PLECS runs into reusable simulation workflows.
Technical scale-upEngineering leadershipAI and data transformation
I work across hardware, software and systems engineering to help teams turn complex products into repeatable, qualified systems. Over 21 years, I have built the tools, products and teams that different fields were missing.
Engineeringteamsdata and AI

Selected work
Simulation tools, engineering systems and experiments with AI. Start with these three, or explore the wider collection.
Start here
Power and systems
Creator and maintainer
A bottom-up model of the AI rack power chain, built to derive the tolerance specifications that do not exist yet.
Creator and maintainer
Modelling a system as a typed component graph so cost, efficiency, availability, and fault trees come from one topology.
Creator and maintainer
A query service over technical standards, built to answer homologation questions against the text.
More selected work
Open-source contributor
Reviewable magnetics models with explicit provenance and honest feasibility boundaries.
Creator and maintainer, built for MenthorQ
Deciding whether a candidate trading feature carries real predictive information, with enough rigour to survive out-of-sample data.
Creator and maintainer
Multi-agent generation and gap analysis of automotive process evidence, with functional-safety cross-referencing.
Across disciplines
2005–2006
CNC and CAM woodwork, pickup winding and frequency-response analysis at a luthier’s bench; then the audio amplifiers that drove them.
Mechanics school, then a year taking instruments apart before any electronics.
2008–2013
FPGA predistortion of GaN power amplifiers; a 21 km laser rangefinder; three patents.
The only field I actually studied. Everything after this was learned on the job.
2013–present
UPS from 500 VA to 1 MVA; the Taycan 800 V three-in-one converter; the OBC7 charger platform — through EMC campaigns, CE and UL.
Never studied it. Learned it by owning a product from schematic to UL certificate.
2016–present
Requirements, architecture, a digital twin in System Composer; test benches from zero, hardware- and software-in-the-loop qualification, ISO 26262 and ASPICE audits.
The converter kept failing at interfaces nobody owned, so I started owning the interfaces.
2021–2025
A system team grown from 7 to about 30, influencing 40; functional safety, cybersecurity and Application Engineering built from zero.
Same method as a converter — find the seam, put a name on it, give it a test.
2026–present
New markets for a 400/800 V converter portfolio — AI data centres, battery storage, industrial power — from a blank sheet, reporting to the CEO.
Markets I don’t come from. Negotiation and homologation turn out to be problem solving with different units.
2021–present
CTO and quant at MenthorQ — gamma, volatility and price-derived signal frameworks running on EC2.
Control-loop stability and options gamma are the same argument. I checked.
2024–present
An MCP server inside pyPLECS, Agents Kindergarten, Standard Expert, and open skills for engineering agents.
Building end-to-end agents instead of talking about them, and publishing what works.
2005–2026
On GitHub since December 2015. My oldest surviving public repository is a PartKeepr fork from March 2017; my earliest non-fork repository is PyAutomaticTestbench, created in September 2018. The AI tooling shown from 2024 is a later part of that history.
One bar per employer or engagement; thickness is the share of a full week, so the second job and the advisory read thin. Overlaps are the point. Click a bar to open the record.
2005
Instrument-maker’s apprenticeship after mechanics school. I started by taking things apart.
2008
Organic RFID, GaN characterization, ultra-low-noise analog, FPGA predistortion — and freelance embedded work from year one.
2010
Laser rangefinders — low-noise analog centimetres from high-power laser drivers. Three patents.
2013
Single- and three-phase UPS from 500 VA to 1 MVA. Product ownership from architecture to UL.
2016
Taycan 800 V three-in-one converter — magnetics, EMC, ISO 26262. Then systems engineering and model-based design.
2021
Head of System, Test and Application Engineering — 7 to about 30 engineers, influencing 40. Quant finance began at night.
2023
Senior quant at MenthorQ, advisor to a semiconductor application team, guest lecturer in power electronics.
2026
Technical business development for AI data centres and storage; engineering advisory; quantitative finance. All active.
Scroll to see more conference photographs.




Experience
Explore my roles, contributions and projects across industry, research and advisory work.
Writing
What I get wrong, what transfers between fields, and where software, data and AI earn their place in physical-product engineering.
Get in touch
Limited hours, so I would rather say that than waste yours. What fits: a technical scale-up assessment, advisory, a second opinion on architecture or verification strategy, a workshop. What does not fit: embedded delivery.
Book an introductory conversationThe tools are open — simulation, magnetics, systems, agents. Fork them, break them, open an issue. If you are stuck on a workflow or a model, I will answer.
Browse the repositories on GitHubAsk me. The writing, the tools and the answers are free. You will be running this industry in twenty years, and I learn as much from these conversations as you do.
Write to me on LinkedIn