Enterprise Generative AI is a self-paced course for professionals who need to plan, evaluate or lead generative AI initiatives inside a business. Eight lessons explain how large language models work and where they fail, how to choose and prioritise use cases, how retrieval-augmented generation and integration patterns work, and how to handle data security, governance, evaluation, cost and adoption. The focus is vendor-neutral and practical.
What you will learn
- Explain how large language models generate output and why they make mistakes
- Identify, score and prioritise enterprise use cases
- Describe retrieval-augmented generation and common integration patterns
- Apply data protection, security and governance controls to AI systems
- Design an evaluation, cost and rollout plan for a pilot
Lessons
- How Large Language Models Work
- Identifying and Prioritising Use Cases
- Prompting and Patterns for Reliable Output
- Retrieval-Augmented Generation and Integration
- Data Protection and Security for AI Systems
- Governance, Policy and Responsible Use
- Evaluation, Monitoring and Cost Management
- Leading Adoption and Change
Self-paced course from Tech Skills Library. Educational content only. Completing the course earns a Certificate of Completion, not an accredited qualification, licence or credential.
Lessons
How Large Language Models Work
Identifying and Prioritising Use Cases
Organisations rarely fail with generative AI because the technology is weak. They fail because they pick the wrong problems, spread effort across dozens of disconnected experiments, or cannot show value. A disciplined approach to use-case selection is the foundation of every successful programme. Where generative AI fits well The technology is strongest where work involves […]
Prompting and Patterns for Reliable Output
In an enterprise system, a prompt is not a casual chat message; it is part of the application's logic. Well-designed prompts make output more consistent, easier to check and safer. Prompting is an engineering discipline, with versioning, testing and review like any other code or configuration. Anatomy of a strong prompt Effective prompts usually contain […]
Retrieval-Augmented Generation and Integration
A general-purpose model knows nothing about your contracts, policies, products or customers. Retraining a model on that information is expensive and quickly becomes out of date. Retrieval-augmented generation, usually shortened to RAG, solves this by finding relevant internal content at the moment of each request and supplying it to the model as context. How RAG […]
Data Protection and Security for AI Systems
Generative AI introduces familiar security concerns in new forms, plus some genuinely new ones. Because these systems process large volumes of potentially sensitive text and can be steered by the content they read, security must be designed in from the start rather than added after a pilot succeeds. Data handling and providers Start by classifying […]
Governance, Policy and Responsible Use
Governance turns good intentions into consistent decisions. Without it, generative AI adoption becomes a patchwork of tools, inconsistent risk judgements and unclear accountability. Good governance is not a barrier; it is what allows an organisation to scale use safely and quickly. An acceptable use policy Every organisation using generative AI needs a clear, short policy […]
Evaluation, Monitoring and Cost Management
A demo that impresses in a meeting is not evidence that a system works. Generative AI output varies with inputs and model versions, so organisations need systematic evaluation before launch and continuous monitoring afterwards, along with clear visibility of cost. Building an evaluation set Create a collection of realistic test cases drawn from the actual […]
Leading Adoption and Change
Technology alone does not create value; people changing how they work does. Many generative AI programmes stall because tools are launched without training, workflows are not redesigned, or employees fear what the change means for them. Leading adoption is a change-management task as much as a technical one. Start with the workflow, not the tool […]