Tim Child

Chief Engineer, Software & AI — Exowatt

Software and AI that have to work in the physical world.

I’m Tim Child. I lead software and AI at Exowatt, where we capture sunlight, store it as heat, and turn it back into electricity on demand. I was the company’s first software engineer.

Before that, I spent six years at the University of British Columbia measuring entropy in nanoscale quantum devices at around twenty millikelvin. Different scale, same job: instruments, infrastructure and software that have to be right, because something real depends on them.

Fig. 1
Charge transition of a quantum dot at several temperatures Four S-shaped curves of dot occupation against gate energy. Warmer curves are broader and their midpoints drift to the right by k T ln 2.
Fig. 1 A quantum dot tuned through a charge transition. Warm it and the step broadens, and drifts sideways by kBT ln 2. That drift is the fingerprint of entropy, kB ln 2 from an electron’s two spin states, showing up in a charge measurement. Measuring signals like this was my PhD.
Now
Chief Engineer of Software & AI, Exowatt
Based
Austin, Texas
Trained
PhD Physics, UBC · BSc Imperial · PGCE UCL
Works in
Python, LLM agents & MCP, AWS, energy systems

§ 01 Now

First software engineer, now Chief Engineer.

I joined Exowatt in June 2025, before it had an in-house software team. Since August 2026 I’ve been Chief Engineer of Software and AI, responsible for the software, AI and digital systems behind a company building clean, dispatchable energy for AI datacenters.

It’s a hardware company, which is the point. Code here ends up controlling equipment, so it gets held to the same standard as the equipment it controls.

  1. 01

    Energy management

    The Exowatt EMS, our energy management system, from architecture through deployment and monitoring.

  2. 02

    AI & agents

    The company’s AI strategy, and the AI tooling and agent infrastructure the team runs on.

  3. 03

    Platform & IT

    Internal apps, IT infrastructure and security, and the engineering partnerships a small team relies on.

§ 02 Path

Tim Child, smiling, in a black t-shirt against a pale grey background
Fig. 2 Tim

Physicist, teacher, engineer. In roughly that order.

  1. 2025 —

    Lead Software Engineer, then Chief Engineer of Software and AI

    Exowatt · Austin, Texas

    Joined as the first software engineer in June 2025. Chief Engineer since August 2026.

  2. 2024 — 25

    Built a generative-AI platform, solo, from the ground up

    AthenaInstruments · independent

    A model-agnostic, Perplexity-style web app for composing agentic workflows over custom retrieval and data pipelines. A Python codebase of around fifty thousand lines, with its own CI/CD and cloud infrastructure.

  3. 2018 — 24

    PhD, Physics — Stewart Blusson Quantum Matter Institute

    University of British Columbia · Vancouver

    Quantum Devices group, supervised by Prof. Joshua Folk. QuEST Scholarship. Thesis on direct entropy measurement in mesoscopic systems.

  4. 2016 — 17

    PGCE in Science, with Qualified Teacher Status

    University College London

    A master's-level teaching qualification, taught in several London schools. Institute of Physics Scholarship.

  5. 2013 — 18

    Private tutor, maths and physics

    London

    Students aged four to eighteen-plus. It helped fund my undergraduate degree.

  6. 2011 — 14

    BSc Physics, First Class Honours

    Imperial College London

§ 03 Research

Measuring entropy, one electron at a time.

Entropy is normally measured in bulk, through quantities like heat capacity. In a device a few hundred nanometres across those quantities all but vanish, so you need a different route. My thesis introduced one: a measurement protocol, built on a Maxwell relation, that works for arbitrary mesoscopic circuits.

I developed it, tested it, and applied it to systems of increasing complexity in GaAs/AlGaAs two-dimensional electron gases: a single quantum dot, a pair of dots coupled so that one could sense the other’s entropy, and a dot coupled strongly enough to its leads to reach towards the Kondo regime, where theory and experiment start to disagree.

Thesis, 2024 Direct Entropy Measurements in Mesoscopic Systems: From proof of concept to the Kondo regime

Selected publications

§ 04 Building

Things I’ve made in the open.

More in the portfolio and the guides, or on GitHub.

How I work

  1. Work out what is most valuable to the people around me, and do that, even when it isn’t the fun part.

  2. AI doesn’t change what engineering discipline looks like. It changes what the discipline costs to maintain.

  3. Physics taught me to care about the details. Teaching taught me to explain them, so they survive the handoff.

§ 05 Contact

Happy to compare notes on energy software, agent infrastructure, or how engineering teams work with AI.

Off the clock Caving, skiing, ice climbing and Arctic expeditions. I once built the hardware and electronics to convert an SUV into a 4×4.