Plain questions, real answers — from your data, your documents, the things your business already knows. I’ve spent thirty years making technology genuinely useful; lately that means AI that gets you straight to what your information is telling you, grounded in what’s actually there rather than made up.
The interesting question is rarely “can we use AI here?” — it’s “should we, and where does it actually hold up?” Sometimes the answer is a language model. Sometimes it’s a clean spreadsheet, a better process, or a small tool that quietly does one job well. I’ve spent thirty years telling the difference — at the top of $4bn enterprises, and lately alongside owners of businesses small enough to feel every decision.
Three projects, one habit — point technology at a real problem and make it behave. Documented end to end, with live output you can explore.
Type a plain question at a real business’s trading data and get a genuine answer, with the chart to match. Built on a garden centre with three years of history — and architected so the AI can only ever answer what the data can actually support.
A working demonstration, going live here soon →Ground-up website, infrastructure and custom tooling for a community men’s shed — built so a non-technical committee keeps it alive without a developer on call. Its bulletin assistant lets anyone ask a plain question of ten years of newsletters and get a grounded, sourced answer.
From a filing cabinet to something you can talk to →A repeatable drone-mapping and photogrammetry pipeline for a vineyard — centimetre-accurate maps, elevation models and 3D point clouds, with a hard cost constraint through every decision.
Explore the live maps and 3D point clouds →The judgement above wasn’t formed in theory. Selected engagements, sponsored directly by the CEO, CFO or owner.
I work with community organisations at rates that suit them, and sometimes at no charge. The Ku-ring-gai Men’s Shed is one — some of the work I’m most glad to have done. If you run a community organisation and technology is getting in your way, get in touch and we’ll find something that works for both of us.
I’m David Eldridge. For thirty years I’ve built and fixed the systems businesses actually run on — in construction, pharmaceuticals, ports, education, finance, healthcare and the charity sector — usually working directly with the person who owns the problem.
The through-line was never a particular technology. It was working out what was actually needed, and making it hold up in daily use. AI is the current chapter of exactly that: an enabler, used where it earns its place.
I come at problems from a lot of angles, because I’ve lived at a lot of angles. I grew up on a remote settlement in the Northern Territory, doing lessons over the radio on School of the Air — either where the self-reliance started or just where I first noticed it. Since then I’ve flown as a pilot, played piano to a level I’m still proud of, built fine furniture by hand, chased light as a travel photographer across close to forty countries, and spent a fair bit of time trying to lower a golf handicap that resists the effort.
The common thread — the one that matters for work — is that I go deep, across unrelated things, and I like getting them right. That turns out to be a useful instinct to bring to someone else’s business.