Deep-Tech Executive | Computer Vision & XR/AR | Real-Time Systems
Everyone in tech is saying some version of “AI makes you 10x.” VCs say it on panels. Founders say it in pitch decks. People selling $49/month productivity tools say it loudest of all. Nobody stops to ask: 10x what, exactly? I keep hearing three different claims, usually mashed together as though they’re the same thing....Continue reading
I get asked about this more often than I expected, so although a post like this will likely age very quickly (as I always upgrade my tech stacks) I thought it would make sense to document my daily AR stack for productivity. I use AR glasses as a portable external monitor that fits in a...Continue reading
People ask me about startups more often than you’d think. Usually some version of: “Why does your company keep raising money? Are you… losing it?” And when they try to look it up themselves, they hit a wall of jargon – cap tables, preferential shares, vesting – that makes the whole thing feel like a...Continue reading
This article reviews two production sports-tracking systems from the early 2000s as practical examples of classic computer vision deployed at scale in complex outdoor conditions. Now that confidentiality constraints have expired, I provide a detailed technical walkthrough focusing on real-time reliability, low-bandwidth distribution, and operating on minimal FLOPS - and how those constraints still inform edge architectures today. The article also draws out lessons for modern edge deployments, where latency, power, cost, and operational risk remain as intertwined as ever.Continue reading
I love demos. I also don’t trust them. Over 25 years I’ve shipped many computer vision (CV) systems that looked great in the lab and became painful the moment they met weather, procurement, and people. This article is about an idea I had in 2012 and shelved. The economics didn’t work. But hardware costs have...Continue reading
I had the good fortune (for a programmer!) to spend years working alongside professional athletes – British association footballers in particular. If there’s one idea that has stayed with me, it’s that measuring performance is a competitive advantage. Not vanity dashboards, but real measurement: what are we trying to improve, how will we know it’s...Continue reading
Updated: 31 January 2026 Much of my career has revolved around growing engineering-led startups into self-sustaining organisations – without losing the thing that made them special: speed. This is a practitioner’s view of the technical transitions that define a company’s first few critical years. The shifting technical landscape In the 1980s and 90s, one-person software...Continue reading
At the end of April 2024 came yet another announcement in autonomous robotics: the Astribot S1. Almost immediately, feeds filled with posts claiming it was nothing short of a technical breakthrough that would transform our kitchens. In this post I explain why I believe it's at best a prototype with little autonomous or genuine real-world capability, and unfit for safe operation around humans.Continue reading
Across over two decades of delivery work, I’ve found that computer vision success is rarely decided by the model alone. This piece captures five lessons that consistently surface when prototypes become products: defining measurable targets, building a trustworthy data/ground-truth pipeline, and making explicit trade-offs across accuracy, latency, compute, and cost. It also highlights common failure patterns - environmental variation, drift, and wrong assumptions - and how to design for them. The emphasis is on repeatable execution and reliability in production.Continue reading