Case study · AI product design and build
RugbyIntel
An AI-powered rugby intelligence app that turns trusted match data into clear, evidence-backed insight for supporters, clubs and organisations. See more than the score.
Three questions after every match
Rugby fans see the score, but rarely the story behind it. RugbyIntel was built as a proof of concept for AI in sport, starting with Premiership Rugby, the United Rugby Championship and the Six Nations. Every screen answers one of three questions.
What happened?
Every match, scored and put in context with verified data, live as it happens.
Why does it matter?
AI explains the significance of a result in plain English, using only checked facts.
What to watch next?
Form, trends and fixtures point you to the matches that will matter most.
Insight, not betting tips. RugbyIntel analyses play, form and performance. It gives no odds and makes no predictions.
Match-day intelligence in your pocket
The app opens straight from a link, with no download. Live scores update on their own, and new insights arrive in the stream as the game unfolds.
Built like a timing screen
The desktop design takes its cues from live sports timing displays: dark, dense and quick to read, with colour reserved for what changes.
The score, and what it means
A live match sits beside its intelligence stream and the effect on the league table, so the context updates with the score.
Every claim can be opened up
Team ratings, recent form and scoring come with a "Why?" panel that shows the working. Where the data is missing, the app says so.
Whole-league trends at a glance
Form leaders, highest scoring teams and the best defences for each competition, alongside fixtures, results and the table.
Facts first, then the AI explains
The app connects to leading AI model APIs from its own secure server. The AI is never asked to guess. The app works out the facts first, and the AI turns them into clear explanations a newer fan can follow.
Trusted match data
Scores, fixtures and tables arrive from a licensed rugby data feed.
Calculations
Form, streaks, scoring and defensive trends are worked out by the app.
AI explains
The facts are sent to an AI model API, which writes the headline and explanation.
Live in the app
The insight is published and appears in the stream within moments.
Invented statistics
The AI may only use the facts it is given. If the facts aren't enough, it says so rather than filling the gap.
Confidence labels
Every card says where its insight comes from, so readers know what is fact and what is interpretation.
Insight triggers
The AI runs only when something worth explaining happens, such as an upset, a lead change or a form streak. That keeps running costs low.
From idea to live AI app
We combine experience of enterprise app delivery with an AI-assisted engineering workflow. That lets us take an idea to a working, connected AI product far faster than a traditional build.
Design
We set out the questions the app had to answer and designed a dark, data-first look for match day.
Build
The app was built with Claude Code, Anthropic's AI coding tool, with an engineer reviewing and directing every change.
Connect
We linked the app to a live rugby data feed and to AI model APIs, with all keys held securely on the server.
Test and host
Automated tests check the calculations before release. The app runs on secure cloud hosting.
Built with Claude Code
Claude Code wrote and refactored the app under our direction, from the screens to the data connections and the rules that keep the AI to the facts. Our engineers set the design, checked the work and made the decisions.
Let's put AI to work for you
We design and build AI apps that connect to your data and explain it clearly. If you have an idea, we can turn it into a working proof of concept.
Screens show demonstration data. Scores, insights and ratings are examples, not real match results. RugbyIntel is not affiliated with any league or club.