Two brothers building the platform for AI vision.
EyeOfTheTiger.ai makes open camera hardware and the software to build on it. Connect any model, build what you need by talking to your agent, and sell it to everyone who owns one.
It started with two unrelated problems and one realisation: they were the same problem.
Charlie and Billy Crawford are brothers. In 2024 they wanted to put automated line-calling into local tennis clubs, the kind of thing Hawk-Eye does for Wimbledon, except the clubs they played at could never dream of affording it. The whole idea was to build a version cheap enough for a normal club to actually buy.
Around the same time, Billy, who’d studied machine learning at Cambridge, got fed up with the baby monitors on the market and built his own at home: a camera, a model, and a few rules about what to alert on. It worked far better than anything he could have bought.
That’s when it clicked. A line judge and a baby monitor are the same machine: a camera, an AI model that understands what it’s looking at, and a handful of rules about what to do next. The hard part was never the app: it was that there was no good camera you could simply point a model at and build on. Every camera was a closed box running someone else’s closed app.
So that became the company. Not a tennis gadget, not a baby monitor: the open platform underneath all of it. A camera with plain MCP tools any agent can call, that runs on your own network, and a marketplace where the apps people build can be sold to everyone else who owns one.
In 2026 Charlie left his job as COO of WeMove to do this full-time. We’re a tiny team, building in the open, convinced that the next wave of useful AI doesn’t live in a chat window: it can see.
Who’s building it
Two founders, one operator, one engineer. Roles still get split down the middle most days.
Charlie Crawford
Spent the last decade scaling operations, most recently as COO of WeMove, the UK’s leading demand-responsive transport operator, where he grew its contract base 10x and launched DrivenConnect and DrivenPeople from scratch. In 2026 he quit to go all-in on AI vision. Charlie runs commercial, the marketplace, and getting the EOT-1 into the world.
Billy Crawford
The technical half. Professional data scientist and machine learning engineer, with a decade of experience building data systems for the UK government. Read maths at the LSE (top of his year, Cyril Offord Prize) and machine learning at Cambridge. He built his own AI baby monitor at home rather than buy one, and that hack became the seed of the EOT-1. Billy owns the platform and device software.
Four things we won’t compromise on
Open by default
No walled garden. Plain MCP, any model, your network. If you can talk to an agent, you can build for the camera, and sell it to everyone who owns one.
Private by design
Vision is intimate. Footage is processed on-device, on your own network, by default. Privacy isn’t a setting you find: it’s the starting point.
Build by talking
The best interface is a conversation. Point your agent at the camera, describe what you want, and ship it. No SDK to learn, no glue code to maintain.
Built in public
Prototypes, dead ends, and milestones, out in the open. We’d rather show the messy middle than a polished launch film.
From a tennis court to a platform
A line call at the tennis club
Hawk-Eye is brilliant, and far too expensive for an ordinary UK tennis club. The brothers start sketching a vision system that ordinary clubs could actually afford.
The baby monitor
Rather than buy an off-the-shelf monitor, Billy builds his own with a camera and a model. It works embarrassingly well, and makes the real insight obvious.
The platform, not the product
Line-calling and baby monitoring were the same problem: a camera plus an AI model plus a few rules. The product isn’t any one app: it’s the camera you point any model at.
All in
Charlie leaves WeMove. EyeOfTheTiger.ai becomes the full-time job: open camera hardware, an MCP layer for any agent, and a marketplace for the apps people build on top.
Batch 001 is in stock and shipping now.
100 Founder Edition units. Ships worldwide. One-time payment, no subscription required, 30-day returns.