Platform · Technical specification
Zia, the skill graph, and adaptive pathways — three pillars under one operating system for workforce capability.
An analyst-report walkthrough of the architecture behind Zetamu: the proprietary AI tutor benchmarked fastest in the 2024 Aragon Research Globe, the largest open skill graph in corporate L&D, and the adaptive algorithm that closes skill gaps in 47% less time than legacy LMS platforms.
Pillar 01 · The AI tutor
Zia is the fastest tutor in the category — and the one learners actually finish.
Zia is Zetamu's proprietary AI tutor. Built on a retrieval-grounded architecture trained against the Zetamu skill graph, she delivers personalised curriculum at a 0.07-second median response time — the figure benchmarked by the 2024 Aragon Research Globe for Corporate Learning as the fastest in the category.
- 97% course-completion rate across 1.84M active learners, vs. an industry average of 14% (Q1 2025)
- 0.07s median response latency, benchmarked fastest in the 2024 Aragon Research Globe
- 38,000 granular competencies grounding every answer, drawn from the Zetamu skill graph
She is not a chatbot bolted on top of a content library. Every reply is anchored to a competency node, recommended against the learner's assessed gap, and re-evaluated after each interaction. The result is a tutor who knows what the learner already knows, what they need next, and — crucially — when to stop talking.
Pillar 02 · The skill graph
The structural layer: 4,200 roles, 38,000 competencies, one open graph.
The Zetamu skill graph is the largest open competency map in corporate L&D. Every role profile — from junior data analyst to AI Risk Officer — is decomposed into the granular capabilities required to perform it, and every learning asset in the catalogue is mapped back to those capabilities at the node level.
The graph is also the first in the category to ship with a SOC-compliant library purpose-built for AI Risk & Governance roles — addressing the audit, model-risk and policy competencies that compliance teams are now formally accountable for. Buyers mapping their workforce against new AI regulation should treat this library as a starting point, not a checklist.
Pillar 03 · Adaptive pathways
The algorithm, in three acts.
Adaptive pathways are where Zia and the skill graph meet. Every learner is benchmarked, sequenced, and re-sequenced against role competencies — closing skill gaps in 47% less time than legacy LMS platforms, measured against the 2024 Verdant Analytics cohort.
-
Act 01 · Benchmark
Every learner is scored against the role, not the catalogue.
On entry, Zia runs a diagnostic pass that maps the learner's existing competencies — drawn from prior roles, project history, and an adaptive diagnostic — to the target role profile in the skill graph. The output is not a course recommendation; it is a gap vector of 30 to 90 specific competencies, ranked by impact on role performance.
- Diagnostic length
- ~18 minutes
- Median competencies scored
- 54
- Refresh cadence
- After every session
-
Act 02 · Sequence
Pathways are re-sequenced in real time, not at enrolment.
The pathway engine re-evaluates the learner's gap vector after every interaction with Zia — adjusting the next module, dropping redundant content, and surfacing stretch material when a competency crosses a confidence threshold. Sequencing decisions are logged and auditable, so L&D leaders can replay exactly why a learner saw a given module on a given day.
- Re-sequence events / week
- 11.4 average
- Audit log retention
- 7 years
- Pathway variants per cohort
- 1:1.7
-
Act 03 · Close
Skill gaps close in 47% less time than legacy LMS.
Across the 2024 Verdant Analytics cohort, learners using Zetamu closed the same competency gap in 47% less calendar time than learners using legacy LMS platforms, holding role profile, prior experience and weekly study time constant. The audit measured 2,140 skill gaps closed per deployment on average, across 312 enterprise rollouts.
- Time-to-competent reduction
- −47%
- Gaps closed / deployment
- 2,140 average
- Audit reference
- Verdant Analytics, 2024
By the numbers · Cohort audit, Q1 2025
The platform isn't a pilot.
- 1.84M Active learners across 312 enterprise deployments
- 312 Enterprise rollouts, audited for the 2024 Verdant Analytics cohort
- 2,140 Average skill gaps closed per enterprise deployment
- 3.4× Average ROI on L&D spend across the 2024 cohort
Procurement checklist
Security, scale, and capital — resolved in one beat.
The questions your vendor-risk assessment will ask, answered in the open. Certifications, named customers, third-party recognition, and capital backing — so the security review doesn't bounce off the page.
-
Certifications
SOC 2 Type II, ISO 27001, and GDPR-certified. Trusted by 47 of the Fortune 500, including Pfizer, Shopify, and Klarna.
Read the security overview → -
Third-party recognition
Named a 2024 Gartner Cool Vendor in Human Capital Management. Ranked #412 on the 2024 Inc. 5000. Winner of the 2024 EdTech Breakthrough Award for Best AI in Education.
See the full proof set → -
Capital & company
Series B closed March 2024 at a $310M valuation: $42M led by Accel, with participation from Emergence Capital and GSV Ventures. Total funding $68M across three rounds. 214 employees across 14 countries as of January 2026.
Company background → -
Customer rating
4.86 / 5 average rating on G2 across 1,204 reviews — the highest in the corporate learning category. 1,840,000+ active learners as of January 2026.
Read verified reviews →
Ready to see the platform against your own competency map?
Book a Capability Assessment