ChatGPT Essentials Training Course

Participants can set up and organise ChatGPT for secure, repeatable professional use.

• ChatGPT plans, model options, and workspace settings compared responsibly
• Configure custom instructions for role, audience, tone, and output preferences
• Start chats, manage history, and organise reusable project materials
• Attach files and interpret supported formats, limits, and handling rules
• Voice, image, and web features matched to appropriate task types
• Create a personal prompt library using names, tags, and examples
• Apply organisational acceptable-use rules before entering business information

Participants can write structured prompts that produce clearer, more usable outputs.

• Anatomy of effective prompts: role, task, context, constraints, and format
• Specify audiences, objectives, source material, and decision criteria precisely
• Use delimiters and structured inputs to separate instructions from reference text
• Few-shot prompting with positive examples, counterexamples, and expected outputs
• Prompt templates designed for repeatable emails, briefs, tables, and analyses
• Request JSON, Markdown, tables, and bullet structures for downstream reuse
• Improve ambiguous prompts through constraint testing and targeted clarification

Participants can manage multi-step ChatGPT conversations and design repeatable task flows.

• Break complex assignments into staged prompts with review checkpoints
• Use follow-up questions to refine scope, assumptions, and recommendations
• Chain drafting, critique, and revision prompts without accepting unsupported claims
• Ask ChatGPT to identify missing information before producing an answer
• Maintain context across sessions using concise handover summaries
• Compare alternative outputs against explicit criteria and stakeholder needs
• Design reusable conversation flows for recurring knowledge-work tasks

Participants can use ChatGPT to draft and improve professional communications under clear review controls.

• Draft stakeholder emails with stated purpose, action, deadline, and tone
• Turn rough notes into meeting agendas, minutes, and action registers
• Produce first drafts of reports using approved outlines and evidence
• Adapt messages for executives, customers, technical teams, and public audiences
• Apply plain-language editing for clarity, brevity, and inclusive wording
• Generate content variations while preserving factual claims and approved terminology
• Use critique prompts to strengthen structure, logic, and reader action

Participants can use ChatGPT to support basic data analysis while independently validating findings.

• Prepare clean tabular data before uploading files for analysis
• Ask for descriptive statistics, trends, outliers, and data-quality questions
• Generate spreadsheet formulas with test cases and formula explanations
• Interpret charts, summaries, and calculations without treating outputs as authoritative
• Build decision tables from operational data, assumptions, and thresholds
• Request Python code, inspect logic, and validate results independently
• Document data sources, transformations, and analytical limitations for reviewers

Participants can extract, compare and summarise information from documents with traceable source checks.

• Summarise long documents using purpose-specific questions and output formats
• Extract obligations, dates, risks, and actions into structured tables
• Compare policy versions and identify material wording differences
• Create research briefs separating supplied evidence from model-generated interpretation
• Use web search results with source checks and citation review
• Build retrieval questions that expose gaps in uploaded reference material
• Respect copyright, confidentiality, and retention requirements for source documents

Participants can test ChatGPT outputs and apply a documented verification process before use.

• Apply a verification workflow: claim, source, check, correction, record
• Detect fabricated citations, calculations, quotations, and policy statements
• Test prompts with edge cases, conflicting inputs, and incomplete data
• Use red-team questioning to reveal unsafe, biased, or weak outputs
• Evaluate output quality using accuracy, relevance, completeness, and traceability criteria
• Define human review thresholds for low-risk and high-impact tasks
• Record assumptions and provenance for material used in decisions

Participants can apply practical controls for sensitive information, unsafe inputs and AI-related risks.

• Classify public, internal, confidential, and restricted information before prompting
• Apply data minimisation, anonymisation, and redaction to sensitive inputs
• Recognise prompt injection attempts in documents, emails, and web content
• Set approval boundaries for customer communications, legal text, and decisions
• Map AI risks using NIST AI RMF Govern, Map, Measure, Manage
• Review third-party terms, retention settings, and enterprise controls
• Escalate suspected data exposure or harmful output through incident procedures

Participants can select a suitable work use case and plan a controlled ChatGPT trial.

• Prioritise candidate use cases using value, feasibility, risk, and ownership
• Define baseline measures for time, quality, error rates, and adoption
• Create standard operating procedures for approved ChatGPT-enabled workflows
• Assign prompt owners, reviewers, and update cycles for reusable assets
• Plan a 30-day workplace trial with controls and success criteria
• Communicate limitations and review responsibilities to managers and colleagues
• Build an individual application plan tied to current work priorities