COMPONENT 01 · COMPETENCY INTELLIGENCE
The model
underneath every
credential.
Break complex domains into measurable capability slices, define proficiency, track growth. Every assessment, lab, and credential runs off one versioned competency model that stays comparable across customers and across years.
Built for curriculum designers · certification bodies · ministries · sponsoring institutions
THE PROBLEM IT SOLVES
Skills frameworks live in slide decks and Word docs. They cannot calibrate.
Hand-crafted competency lists drift between authors and over time.
There is no defensible link from a credential back to the evidence underneath.
Cross-customer comparability vanishes the moment two cohorts use slightly different rubrics.
A credential is only as good as the model behind it. Competency Intelligence is that model — versioned, auditable, and shared across every Ascend deployment.
HOW IT WORKS
Define → Slice → Sponsor → Calibrate.
Four moments in Competency Intelligence — the surfaces ministers, SMEs, and reviewers actually touch.
Map the capability tree.
Every domain decomposes into Competencies, Levels, and Slices — the atomic testable intents linking a question on screen to the framework it lives under. Shipping today: Agentic DevOps and English Workplace (CEFR A1–A2).
- L1 / L2 / L3 levels map to industry standards where available, e.g. CEFR
- Every layer is versioned — what you trust today, you can still compare against in 2030
SMEs contribute. Calibration stays platform-owned.
Sponsors gain authoring rights inside a scoped competency set: authors draft, reviewers approve, Leanovated calibrators run the final statistical check. Sponsors get the depth; comparability holds across every other sponsor.
- Five-role lifecycle: Author → Reviewer → Calibrator → Joint publish → Monitor
- Sponsors never see each other's items — calibration is platform-side only
Sponsored scope · Harambee × English Workplace
12 items in draft · 8 in calibration · 5 published
5
Harambee SMEs
2
Sponsor side
Leanovated
Final sign-off
The item bank is the framework, made testable.
Every item maps to exactly one Slice and is versioned alongside its framework. That mapping is what makes credentials auditable — and what makes cross-cohort, cross-sponsor analytics possible.
- Trust tier (Bronze / Silver / Gold) recorded per item, set by calibration
- Bias and discrimination checks run before any item reaches Gold
Item bank
English Workplace · v0.1 · 15 items
| ID | Slice | Type | Level | Max tier |
|---|---|---|---|---|
| ITEM-0008 | D3.2 | SPK | L2 | Gold |
| ITEM-0010 | D4.2 | SWR | L2 | Gold |
| ITEM-0013 | D5.2 | SWR | L2 | Gold |
| ITEM-0001 | D1.1 | MC | L1 | Silver |
| ITEM-0004 | D2.1 | MC | L1 | Silver |
Version every change. Defend every credential.
The framework evolves every 6 to 12 months — and every change is recorded. Credentials are stamped with the version they were earned against. No silent updates, no eroded trust.
- Side-by-side version diff, with the calibration cohort attached to each release
- Retired items annotated so historical credentials stay defensible
Framework · Agentic DevOps
v1.1 → v1.2 · 12 May 2026 · Published
WHO USES IT
One model. Many authors. Same calibration underneath.
Universities & TVET institutions
Map graduate readiness against an industry-recognised framework employers can actually read.
Sponsoring institutions
Contribute items inside a scoped competency. Co-brand the framework. Cross-sponsor calibration preserves the value of the credential.
Ministries & certification bodies
A national or sectoral credential infrastructure built on a model you can audit, evolve, and defend.
GET STARTED
Start with one framework. Compose into the platform.
Most customers begin by sponsoring or adopting one framework — English Workplace, Agentic DevOps, or your own. Reach out and we'll walk through the authoring lifecycle on your domain.