Skip to content
Back to work
Football match analyticsAI app2026

AIsvenskan

AIsvenskan scans every Allsvenskan match to find the bets actually worth making. It surfaces value plays with clear reasoning and confidence, tracks its record in the open, and, just as importantly, tells you when there is no edge worth taking.

Year
2026
Status
Live
Stack
ReactSupabaseClaudeChatGPTLovable
What I built
Product concept, how bets and confidence are presented, and the UX for exploring the best plays.
AIsvenskan

How it works

  1. 01

    Every prediction carries the model version that produced it, so a forecast can be traced back to the thing that made it rather than to the app in general.

  2. 02

    Runs are scheduled and the next one is published on the page, alongside how old the result and odds data are.

  3. 03

    The accuracy record is published in full, settled bets and all, with the sample size next to it and a plain statement when the sample is still too small to conclude anything. The log is hash chained, and it says so on the page when the chain breaks.

  4. 04

    It only flags a value signal when the odds diverge clearly from the model, so most days it recommends nothing at all, which costs engagement and is the point.

  • Decision 1

    Surface the best bets, not just raw predictions.

  • Decision 2

    Show confidence and reasoning openly, no black-box hype.

  • Decision 3

    Cover the whole slate so nothing good is missed.

The public ledger, in Swedish. Every settled bet stays on the page, and the page itself says the sample is too small to conclude anything.
The public ledger, in Swedish. Every settled bet stays on the page, and the page itself says the sample is too small to conclude anything.
The training ground: which model is in service, how the next one is built, and the scheduled test that decides whether the current one is kept.
The training ground: which model is in service, how the next one is built, and the scheduled test that decides whether the current one is kept.

Happy to walk through the thinking, the trade-offs and where it's headed next.

Get in touch