Public sector · GCC
Gulf Government Training Initiative — AI training
Delivered a hands-on AI enablement program for 200+ public-sector staff, covering prompt engineering, use-case discovery, and responsible adoption.
Timeline
16 weeks · 4 cohorts
Team
2 lead instructors, 1 curriculum designer, 2 lab engineers, 1 program manager
Focus
AI training
The problem
- Leadership had a top-down AI mandate but no shared vocabulary, success criteria, or delivery playbook across departments.
- Staff were experimenting with public chatbots, creating governance and data-leakage risk.
- Previous vendor trainings were lecture-heavy and produced no internal capability.
The solution
- Role-based curriculum: executives, product owners, analysts, and engineers each followed a tailored 4-week track.
- Hands-on labs in a sandboxed AI environment with their own (anonymized) data — every participant shipped a working prototype.
- Use-case discovery workshops that produced a scored, deduplicated backlog aligned to strategic priorities.
- A lightweight internal playbook covering data handling, evaluation, vendor selection, and a go/no-go gate for production AI.
Architecture highlights
Learning platform
Self-hosted LMS with SSO; lab access scoped per cohort with automatic teardown.
Sandbox
Isolated AI workbench with EU/GCC-hosted models, prompt logging, and red-team test harness.
Governance
Internal risk taxonomy, model registry template, and approval workflow co-designed with legal and IT.
Measurement
Pre/post skill assessments, prototype demos, and a quarterly use-case portfolio review.
Self-hosted LMSSandboxed AI workbenchRegion-hosted LLMsUse-case scoring framework
Measurable outcomes
Indicative figures from the engagement. Exact numbers are placeholders pending client approval to publish.
200+
Staff certified
Across 4 cohorts and 9 departments.
37
Prototypes shipped
By participants during the program.
1
Internal AI playbook
Adopted as the standard go/no-go gate.
Working on something similar?
Let's scope a pilot or a discovery sprint.