Programme overview
Technical depth: Practitioner · Practical mode: Lab
Introduction
Most organisations have given staff access to AI assistants, yet use stays uneven: a few people save hours each week while others get vague answers, paste confidential data or stop trying. The difference is rarely the tool; it is how tasks are framed, checked and shared. This Core Concept course gives teams a structured way to write prompts, apply AI to everyday document and analysis work, and verify what it produces. Participants work hands on in common enterprise assistants and leave with a Team Prompt Library and AI Workflow Playbook built around their own recurring tasks.
Course Objectives
- Explain how large language models produce output and where their limits create risk for team work
- Write structured prompts using the CO-STAR framework, few-shot examples and step-by-step decomposition to obtain consistent results
- Apply AI assistants to summarising, drafting, analysis and document question answering within the team's acceptable use rules
- Evaluate AI output for accuracy, bias and confidentiality before it is used or shared
- Configure reusable custom assistants and prompt templates that other team members can run
- Produce a Team Prompt Library and AI Workflow Playbook with a measured estimate of time saved
Target Audience
- Team members who produce reports, correspondence and briefing documents
- Administrative and executive support staff managing meetings, schedules and communications
- Analysts who summarise data, research and documents for decision makers
- HR, finance and procurement officers handling repetitive document-based work
- Team leaders responsible for standardising how their team uses AI assistants
- Customer and stakeholder communication officers drafting responses at volume
Course Outline
Day 1: Generative AI at Work: Foundations and Ground Rules
- How Large Language Models Generate Text: Tokens, Context Windows and Probabilities
- Enterprise Assistant Landscape: Microsoft 365 Copilot, ChatGPT Enterprise, Gemini and Claude
- Acceptable Use Policy and Data Classification Rules for AI Tools
- Team Task Inventory and AI Suitability Matrix
- Time-and-Quality Baseline Measurement for Target Tasks
Day 2: Prompt Frameworks and Techniques
- CO-STAR Prompt Framework: Context, Objective, Style, Tone, Audience and Response
- Role, Task and Format Prompt Pattern
- Zero-Shot, Few-Shot and Example-Driven Prompting
- Chain-of-Thought and Step-by-Step Decomposition Prompts
- Structured Output Prompts for Tables, JSON and Templates
Day 3: Everyday Productivity Applications
- Document Summarisation and Meeting Minutes Workflows
- Email, Report and Policy Drafting with Iterative Refinement
- Spreadsheet Analysis and Formula Generation Prompts
- Grounded Question Answering over Uploaded Documents
- Custom Assistant Configuration: GPTs, Copilot Agents and Gems
Day 4: Output Quality, Risk and Optimisation
- Confabulation Risk Controls from NIST AI 600-1 and a Source Verification Checklist
- OWASP Top 10 for LLM Applications: Prompt Injection and Sensitive Information Disclosure
- Output Evaluation Rubric for Accuracy, Bias and Tone
- Prompt Iteration Log and A/B Comparison of Prompt Variants
- Redaction and Anonymisation Techniques Before Prompting
Day 5: Lab Practice and the Team Prompt Library
- Function-Specific Lab: Finance, HR and Operations Prompt Scenarios
- End-to-End Workflow Build: From Request Intake to Approved Output
- Prompt Library Structure, Naming and Version Control
- Team Prompt Library and AI Workflow Playbook Drafting
- Peer Testing and Productivity Gain Estimate Against Baseline
Skills You Will Gain
- Structured Prompt Writing
- Few-Shot Example Design
- AI-Assisted Drafting
- Document Summarisation
- AI Output Verification
- Data Confidentiality Awareness
- Prompt Library Management
- Workflow Design with AI
Why Attend This Course
- Return with a Team Prompt Library and AI Workflow Playbook built on your own tasks and tested by peers
- Shift time from first drafts, summaries and minutes towards the judgement work that only people can do
- Know which information must never go into an AI tool and how to check output before it leaves the team
- Practise on assistants that organisations commonly license, alongside participants from different sectors
Conclusion
AI assistants raise team productivity only when prompting, checking and sharing become consistent habits rather than individual experiments. This course moves from how language models work and where they fail, through recognised prompt frameworks and techniques, to hands-on use in drafting, summarising and analysis, and to the verification and confidentiality checks that keep output reliable. The final day turns that practice into a Team Prompt Library and AI Workflow Playbook, giving each team a shared, tested starting point for using AI in its daily work.