Skip to content
Start Free →
THE SCAFFOLD

The diagnostic engine that shows where your judgment breaks.

Bayesian Knowledge Tracing. Graph-native. Issues verifiable credentials.

Run your first diagnostic — free

Live now: Try the Scaffold diagnostic at app.cubelet.ai →

$400B+

spent on training annually

74%

of senior leaders say companies lack skills

Most of that $400B targets guesses, not diagnosed gaps. The Scaffold changes that.

A Bayesian model that updates its estimate of your competency with every assessment response. Not a quiz. Not a self-assessment. A calibrated diagnostic.

HOW THE SCAFFOLD WORKS
01

Scaffold reads your role + domain

The diagnostic engine loads your certification domain and role context from the knowledge graph. No guesswork — structured assessment from first response.

02

BKT runs, HAS_GAP is written

Bayesian Knowledge Tracing updates your competency estimate with every response. When mastery drops below 0.60, HAS_GAP is written to Neo4j.

03

Mirror Moment with credential

Your first gap report, in plain language. The Scaffold issues an issue_diagnostic_vc as a W3C VC 2.0 verifiable credential.

TECHNICAL ARCHITECTURE
Knowledge Graph Neo4j

Unified vector + graph — semantic search and graph traversal in a single substrate.

Mastery Threshold BKT < 0.60

Below threshold triggers HAS_GAP. APPLY-face assessment triggers HAS_MASTERED.

Credential Output issue_diagnostic_vc

W3C VC 2.0 + SD-JWT. Selective disclosure — show pass/fail without revealing scores.

White Space No competitor

No competitor has a Bayesian-calibrated gap diagnostic connected to a content production engine.

FREQUENTLY ASKED QUESTIONS
What is Bayesian Knowledge Tracing?
BKT is a probabilistic model that estimates your mastery of each competency domain. Unlike static tests, it updates its assessment with every response — identifying exactly where your judgment breaks down.
How accurate are the diagnostics?
The Scaffold uses calibrated Bayesian inference across multiple assessment responses. Accuracy improves with each interaction. The 0.60 mastery threshold is set to minimize both false positives and false negatives.
What is a verifiable credential?
A W3C VC 2.0 digital credential that cryptographically proves your diagnostic results. Readable by Microsoft Entra, Apple Wallet, and Google Wallet. Selective disclosure via SD-JWT lets you share only what is relevant.
Can the Scaffold work standalone?
Yes. The Scaffold runs independently as a diagnostic engine. It also integrates with Cubelets (for gap remediation) and the Passport (for credential issuance) as part of the full GRID42 platform.

Standalone or composable

Use the Scaffold alone for gap diagnostics, or connect it to the Cubelets. When a gap is diagnosed, Course Factory routes targeted content automatically. Both compound.

Start free →