Academy / Courses Course catalogue · Available soon

One core, three sector tracks

Every practitioner starts with GovKit Core: how to deliver AI with governance built into the work. Then each person takes the track for their sector. Leaders can join each track through a half-day executive briefing.

Start here GovKit Core Governed AI delivery · 3 days

Track GAI in government services4 days · capstone

Track BAI in banking under regulation4 days · capstone

Track TAI in telecoms operations4 days · capstone

Finish Sector capstone A governed AI release plan for a real case, reviewed by a panel

same core, then your own regulator and frameworks

1Learning path Practitioner route · executives can join each track at the briefing

01 · Catalogue

Four courses open for registration of interest

Each course has a practitioner bootcamp. Each sector track also has a half-day executive briefing. We publish cohort dates when a course is released.

Course	For	Format	Planned length	Prerequisite	Status
GovKit Core	Architects, engineers, delivery and risk leads	Bootcamp with graded lab	3 days, or 6 online half-days	None	Available soon
AI in government services	Government entities and GovTech teams	Bootcamp, capstone, executive briefing	4 days + 1-day capstone	GovKit Core	Available soon
AI in banking under regulation	Banks, finance houses and their technology teams	Bootcamp, capstone, executive briefing	4 days + 1-day capstone	GovKit Core	Available soon
AI in telecoms operations	Operators and their network and IT teams	Bootcamp, capstone, executive briefing	4 days + 1-day capstone	GovKit Core	Available soon

Planned outlines. Modules, lengths and dates can change before release. Classroom, online and hybrid delivery for each course.

Core · all sectors

GovKit Core

Deliver AI features with governance built into the work, not added after it. Specs, architecture contracts, evaluation gates and evidence live in the same repository as the code.

For: Architects, engineers, product owners, delivery and risk leads
Format: Bootcamp with one graded lab · classroom, online or hybrid
Length: 3 days, or 6 online half-days
Prerequisite: None. Basic Git is useful.
Evidence: Graded lab, TILT assessment, lab receipt
Status: Available soon

ArchiMate 3.2ADRsGherkinLLM evaluationISO/IEC 42001NIST AI RMF

Modules

No.	Module	After this module you can
C1	Why governance belongs in delivery	Explain how a control framework (ISO/IEC 42001, NIST AI RMF, EU AI Act) becomes checks in a delivery pipeline.
C2	Architecture contracts	Write layer rules and an ArchiMate view that an AI coding agent must follow. Record decisions as ADRs.
C3	Specs before code	Write acceptance criteria in Gherkin, non-functional requirements and evaluation criteria for one AI feature.
C4	Evaluation gates for LLM features	Set evaluation thresholds and connect them to a CI quality gate that blocks a failed release.
C5	Approval and evidence	Configure named approvers for ADRs. Export evidence that an auditor can trace from requirement to release.
C6	Run it in your organisation	Choose a maturity level and a rollout plan. Assign the roles that own each gate.

Graded lab

Govern one AI feature from spec to release

In a sandbox repository with synthetic data, you write the spec and the evaluation criteria, pass the quality gate, and get your ADR approved by a named reviewer. A deterministic grader checks the result.

Labs use the open-source govkit toolkit (Apache 2.0) for governed AI delivery with Claude Code, Copilot or Codex. The course teaches the method, so the controls apply with other tools too.

Track G · Government

AI in government services

Design and release AI services that citizens can trust. Keep data sovereign, decisions explainable and people accountable for each decision.

For: Federal and local government entities, GovTech and shared-service teams
Format: Bootcamp, capstone and a half-day executive briefing
Length: 4 days + 1-day capstone
Prerequisite: GovKit Core (practitioners)
Status: Available soon

TOGAFArchiMate 3.2BPMN · DMN · CMMNData protectionWardley maps

Modules

No.	Module	After this module you can
G1	Public value and citizen trust	Select AI use cases by public value. Write the transparency notice a citizen sees.
G2	Data sharing and sovereignty	Classify data and apply GCC data-protection laws, for example the UAE and Saudi PDPL, to a data-sharing case.
G3	Architecture for shared platforms	Model a cross-entity capability map in ArchiMate. Use the TOGAF ADM to plan a shared AI platform.
G4	Procure and assure AI	Write AI requirements for a tender. Evaluate vendor models with model cards and acceptance tests.
G5	Agentic services with human decisions	Separate process, decision and case work with BPMN, DMN and CMMN. Place the human decision in DMN.

Capstone

Redesign one citizen service

Teams add an AI agent to a real service journey, with a decision model, a data-sharing agreement and a governed release plan. A panel reviews each plan.

Executive briefing · half day

Decide where AI serves the public

For directors-general and executive directors. A Wardley map of the service landscape and three decisions to take.

Track B · Banking

AI in banking under regulation

Take AI use cases from idea to production under model-risk and supervisory rules. Map each use case to BIAN, so architecture, risk and business speak one language.

For: Banks, finance companies, and their architecture, data, risk and engineering teams
Format: Bootcamp, capstone and a half-day executive briefing
Length: 4 days + 1-day capstone
Prerequisite: GovKit Core (practitioners)
Status: Available soon

BIANModel riskCBUAESAMAArchiMate 3.2

Modules

No.	Module	After this module you can
B1	BIAN for AI use cases	Map an AI use case to BIAN service domains and their control records.
B2	Model risk management	Build a model inventory entry, a validation plan and a monitoring plan against supervisory standards such as the CBUAE Model Management Standards.
B3	Credit, fraud and financial-crime models	Test a model for explainability and fairness. Write the reason codes a customer receives.
B4	Conversational and agentic banking	Define bounded agency for an agent: mandate, consent, limits and a revocation path.
B5	Resilience, cloud and third parties	Assess cloud and AI-vendor risk with outsourcing rules, using DORA as a reference model.

Capstone

From BIAN map to model-risk dossier

Teams take one AI use case from its BIAN mapping to a dossier that a model-risk committee can approve. Labs use synthetic banking data only.

Executive briefing · half day

Govern AI as a board

For ExCo and board members. Model risk appetite, the regulator's view and the questions to ask management.

Track T · Telco

AI in telecoms operations

Put AI into network, service and customer operations with TM Forum ODA, and keep the legacy Frameworx map that most operators still run on.

For: Operators and their network, IT, data and customer-operations teams
Format: Bootcamp, capstone and a half-day executive briefing
Length: 4 days + 1-day capstone
Prerequisite: GovKit Core (practitioners)
Status: Available soon

TM Forum ODAeTOMSIDTAMAIOps

Modules

No.	Module	After this module you can
T1	ODA components and Open APIs	Place an AI capability in an ODA component and name the Open APIs it exposes and uses.
T2	eTOM, SID and TAM	Show which processes, data entities and applications an AI use case changes.
T3	AIOps and autonomous networks	Assess a network domain against the TM Forum Autonomous Networks levels. Plan the next level.
T4	Customer AI	Design a care agent or next-best-action model that respects consent and complaint rules.
T5	Data, sovereignty and regulators	Apply residency and lawful-access rules from regulators such as TDRA and CST to an AI data flow.

Capstone

Design an ODA-aligned AI component

Teams specify one AI component with its ODA placement, SID data, eTOM process change and governance gates. A panel reviews each design.

Executive briefing · half day

Choose the next autonomy level

For CTO, CIO and CXO teams. Where AI pays back in operations, and the risks to accept or refuse.

02 · Register interest

Reserve a place in the first cohorts

Tell us which course and how many people. We confirm dates when the course is released. Private cohorts for one organisation are also possible.

Name the course, the track and the number of people.

Confirm

We send the cohort dates and the pre-work when the course is released.

Prepare

Pre-work opens two weeks before day one. Labs run in a sandbox with synthetic data.