● AI talent-ID engine

The AI that finds hidden talent

Talent Discovery AI reads verified performance, skill and fitness signals across regions and stages to surface athletes who would otherwise go unseen — with explainable scores, explicit confidence, and a human in the loop. It is guidance to act on, not a verdict on a career.

The intelligence layer

How the engine works

Four models, one job: turn verified athlete data into honest, actionable signals. Where an output is an estimate, we say so — and we show the confidence behind it.

AS

AI Scout Live

Surfaces emerging and hidden talent across regions, sports and stages — ranked candidates a human still chooses to pursue.

CG

AI Career GPS Live

Maps current level → next milestone → the gap between them → a concrete development plan, computed per athlete.

SP

Selection Predictor Estimate

A probability of selection with an explicit confidence band. It is a modelled estimate from limited data — never a promise of selection.

VA

AI Video Analysis Soon

Technique scoring and auto-highlights from match footage. In development — early outputs will ship clearly flagged as experimental.

The inputs

What it measures

The engine is only as honest as its data. Every signal below is drawn from verified records — and weighted by how complete and recent that data is.

PS

Performance signals

The measurable outputs of how an athlete actually competes, tracked over time rather than from a single standout day.

  • Event & match results history
  • Trend & trajectory, not snapshots
  • Recency-weighted scoring
SB

Skill baselines

Sport-specific technical and tactical markers that describe what an athlete can do, not just what they scored.

  • Position & discipline-aware metrics
  • Test-based skill assessments
  • Coach-recorded observations
FM

Fitness markers

Physical capacity indicators that put performance in context and flag development headroom.

  • Speed, power & endurance tests
  • Age- & stage-adjusted norms
  • Growth-aware for young athletes
MC

Match context

The same numbers mean different things in different settings — so context is part of every signal.

  • Level & strength of opposition
  • Conditions & competition tier
  • Minutes & role played
PB

Peer benchmarking

The Talent Index places an athlete in context against a relevant peer group rather than an abstract scale.

  • Same age, sport & region cohorts
  • Percentile-style positioning
  • Confidence-flagged composites
VP

Verified provenance

Signals only count when their source can be trusted — the Verification Network proves where each record came from.

  • ed25519-signed achievements
  • Tamper-evident history
  • Source weighting by trust score
Our commitment

Honest by design

An AI that decides careers would be reckless. Ours is built to inform the people who do — transparently, with its limits in plain sight.

CF

Confidence-flagged outputs

Every estimate ships with a confidence indicator. Thin or stale data lowers it visibly, so no one mistakes a guess for a fact.

  • Explicit confidence on estimates
  • Data-completeness signalling
  • Estimate vs. verified, always labelled
EX

Explainable scoring

No black boxes. Each score can be traced to the signals that drove it, so athletes and scouts can see the why.

  • Drivers shown behind each score
  • Sport-specific, documented models
  • Auditable, reproducible results
HL

Human-in-the-loop

AI suggests; humans decide. The engine ranks, flags and explains — coaches, scouts and selectors make the call.

  • Recommendations, never autonomous picks
  • Human-gated nudges & outreach
  • Minor-safe by default

Discover talent others miss

Put the talent-ID engine to work — surface the athletes, then make the call yourself.