Enterprise Generative AI Adoption Training Course

Participants can identify, assess and rank generative AI use cases using explicit decision criteria.

• Use-case canvases capturing users, decisions, workflows, data, and expected outcomes
• Jobs-to-be-Done analysis for employee, customer, and operational pain points
• Value-effort-risk scoring matrix for initial opportunity screening
• Prioritisation criteria covering benefit, feasibility, control maturity, and strategic importance
• Process decomposition identifies task steps suitable for augmentation or automation
• Dependency mapping exposes data, integration, policy, and ownership constraints

Participants can define the architectural and data requirements for enterprise generative AI services.

• Reference architecture: user interface, orchestration, models, retrieval, tools, and monitoring
• Retrieval-augmented generation design using approved enterprise knowledge sources
• Data classification rules for public, internal, confidential, and restricted information
• Identity and access management controls for prompts, documents, and outputs
• API integration patterns connecting generative AI services to business systems
• Model selection criteria covering capability, latency, context limits, hosting, and cost

Participants can identify generative AI risks and specify proportionate preventive and detective controls.

• NIST AI RMF functions: Govern, Map, Measure, and Manage
• OWASP LLM Top 10 threats and mitigation approaches
• Prompt injection defences using instruction hierarchy, isolation, and tool permissions
• Output controls for hallucination, harmful content, bias, and unsupported claims
• Threat modelling identifies assets, attack paths, controls, and residual risk
• Evaluation datasets test accuracy, safety, security, and domain-specific performance

Participants can establish governance structures, policies and approval routes for generative AI adoption.

• ISO/IEC 42001 management-system concepts applied to AI accountability
• RACI matrices assign ownership for use cases, models, data, and controls
• Risk-tiering criteria distinguish low, moderate, high, and prohibited uses
• AI policy clauses address acceptable use, data handling, disclosure, and monitoring
• Approval workflows connect business cases, risk reviews, architecture, and legal decisions
• Assurance evidence includes testing records, decisions, exceptions, and incident logs

Participants can design a practical operating model for enterprise generative AI services.

• Centralised, federated, and hybrid AI operating-model archetypes compared
• Product operating model defines persistent teams, backlogs, funding, and accountability
• Shared AI platform services cover model access, guardrails, evaluation, and support
• Communities of practice coordinate standards, reusable assets, and practitioner learning
• Service catalogue defines approved models, capabilities, support levels, and pricing
• Control ownership separates first-line delivery from independent oversight functions

Participants can evaluate sourcing options and manage third-party generative AI dependencies.

• Build-buy-partner analysis compares differentiation, speed, capability, control, and cost
• Total cost of ownership includes licences, integration, monitoring, support, and change
• Vendor due diligence assesses data use, model hosting, security, resilience, and exit
• Contract clauses address intellectual property, audit rights, incident notification, and liability
• Model provider concentration risk assessed through substitution and contingency options
• Proof-of-value criteria establish evidence before procurement scale commitments

Participants can apply delivery stages and quality controls from discovery through live operation.

• Stage-gate delivery model from discovery through retirement and decommissioning
• Product requirements specify users, tasks, constraints, failure modes, and acceptance criteria
• Human-in-the-loop design sets review thresholds, authority, and override procedures
• Red-team testing probes misuse, jailbreaks, data leakage, and unsafe tool actions
• Model cards document intended use, limitations, evaluation results, and known risks
• Production monitoring tracks quality drift, incidents, usage patterns, and costs

Participants can measure adoption outcomes and plan workforce change for generative AI services.

• Benefits-realisation maps connect AI outputs to operational and financial measures
• Baseline measures capture cycle time, quality, cost, satisfaction, and risk
• Balanced scorecards track value, adoption, control performance, and technical reliability
• Change impact assessment identifies altered tasks, roles, skills, and decision rights
• Training pathways distinguish general users, supervisors, developers, and risk owners
• Communication plans explain approved uses, limitations, escalation channels, and accountability

Participants can assemble a sequenced adoption roadmap with owners, controls, investments and measures.

• Roadmap sequencing balances quick wins, platform foundations, controls, and capability building
• Ninety-day plan assigns actions, owners, dependencies, milestones, and decisions
• Funding models compare project budgets, product funding, and shared-service charging
• Executive dashboard presents portfolio value, risk status, adoption, and delivery progress
• Decision log records assumptions, approvals, exceptions, and accountable decision-makers
• Application plan identifies immediate actions for participants' organisational priorities