Learning and careers decision
Refereegpt
A single technical user can implement the core personal trainer (scenario input, LLM answers with citations, quiz state) in a few weeks, but reproducing the product's team/federation features, API access, and white‑label/enterprise capabilities would require more work and operational support.
Visit website↗Built by Kevin Middleton | FootballGPT, who ships 7 products in this index
$9.99/mo
$120/yr
Read off the official pricing page.
$100one-off50 h to build
$70/mo3 h/mo upkeep
On cash alone, building overtakes the subscription at 8 seats.
No open-source build does this yet
Nothing published replaces this one, so a replacement starts from an empty file. Here is what it would have to cover.
What a replacement has to do
- Accept an exact match scenario from the user, call an LLM with the scenario plus the applicable ruleset, return a decision with reasoning and a cited official rule, record the attempt and update quiz/heart/streak state.
What it still won’t have
- Association/Federation features (bulk invites, admin dashboard)
- API access / white‑label / custom branding
- Priority support and dedicated account manager
- Built-in team seat management and org analytics
What remains hard
- Product polish and ongoing maintenance
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 8 seats.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a single‑user Referee training web app using Next.js (React) + Node API, Postgres for data, Vercel for hosting, OpenAI-compatible LLM API for reasoning, and Stripe for subscriptions. In scope: user signup/login, scenario input form, prompt pipeline that attaches the selected sport/ruleset, store and serve rule citations, AI response display with decision + explanation + citation, quiz mechanics (hearts, streaks, history), subscription gating for unlimited questions, basic analytics (per-user attempt count). Out of scope: multi-seat org admin features, white‑labeling, public API, advanced analytics dashboards. Include input validation, error handling for API failures and rate limits, unit and integration tests for core flows, and deployment scripts.
How we checked
How the score was reached
- Partly verdict base52
- 2 cited sources+1
- Price verified on pricing page+3
- 3/3 assessment runs agreed+4
- Evidence score60
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time - so the same evidence always produces the same number.
How scoring works →Cited sources · 2
Every page the run actually retrieved.
- official productReferee Training for Match Scenarios | RefereeGPT
- official pricingPricing - Plans for Referees & Associations | RefereeGPT
Integrity checks
What held up, and what did not.
