AI Prompt Engineering Course

Lesson 1 • The Prompt-Response Relationship

Defines the input-output loop between user prompts and model responses. Shows how context window size and message structure shape what the model can produce.

Lesson 2 • Key Terminology and Mental Models

Defines hallucination, grounding, inference, and fine-tuning in practical terms. Provides shared vocabulary used throughout the entire course.

Lesson 3 • How Large Language Models Work

Covers token prediction, training data, and probability distributions in plain terms. Establishes the mechanical basis that explains why prompt wording changes outputs.

Lesson 1 • Instruction Verbs and Action Words

Catalogues high-signal verbs such as summarise, classify, extract, and rewrite and their effects. Demonstrates how verb choice directly controls the type of output generated.

Lesson 2 • Clarity, Specificity, and Conciseness

Teaches how vague language produces vague outputs and how precise wording narrows model behaviour. Applies editing techniques to transform weak prompts into strong ones.

Lesson 3 • Common Structural Mistakes to Avoid

Identifies over-long preambles, contradictory instructions, and missing output specs as frequent failure modes. Provides a checklist for self-auditing prompt structure before submission.

Lesson 4 • Formatting Techniques for Better Output

Covers markdown, delimiters, numbered lists, and XML-style tags as structural tools. Shows how formatting signals intent and organises complex instructions.

Lesson 1 • Few-Shot Prompting with Examples

Teaches how to embed multiple input-output examples to guide model behaviour towards a target pattern. Covers example quality, quantity, and ordering for maximum effect.

Lesson 2 • Zero-Shot and One-Shot Prompting

Defines zero-shot prompting as instruction without examples and one-shot as instruction with a single example. Establishes the baseline from which all other techniques build.

Lesson 3 • Constraint and Negative Prompting

Uses explicit exclusions and boundary conditions to prevent unwanted content or format. Pairs positive instructions with negative constraints for precise output control.

Lesson 4 • Role and Persona Prompting

Demonstrates how assigning a role or persona shifts tone, vocabulary, and reasoning style. Covers appropriate use cases and risks of persona-based framing.

Lesson 1 • Systematic Prompt Testing Methods

Introduces controlled variation testing where one prompt element changes per iteration. Builds habits of evidence-based refinement rather than random rewrites.

Lesson 2 • Diagnosing Poor Model Outputs

Teaches a root-cause framework for identifying whether failures stem from ambiguity, missing context, or wrong technique. Connects diagnosis directly to targeted fixes.

Lesson 3 • Using Model Feedback Loops

Shows how to ask the model to critique its own output and suggest prompt improvements. Leverages self-evaluation as a low-cost refinement accelerator.

Lesson 1 • Question Answering and Research Tasks

Structures prompts for factual Q&A, comparative analysis, and research synthesis tasks. Teaches grounding techniques to reduce hallucination in knowledge-intensive outputs.

Lesson 2 • Information Extraction Prompts

Demonstrates structured extraction of entities, dates, relationships, and key fields from unstructured text. Uses output format specifications to ensure machine-readable results.

Lesson 3 • Summarisation and Condensation Prompts

Covers length targets, audience framing, and key-point preservation for effective summarisation. Addresses common failure modes such as hallucinated details and omitted critical content.

Lesson 4 • Content Generation and Creative Tasks

Applies tone, style, audience, and format constraints to guide open-ended generation tasks. Balances creative freedom with output consistency through layered constraints.

Lesson 1 • Designing Multi-Turn Conversations

Teaches how to maintain context, manage topic shifts, and preserve instruction adherence across multiple exchanges. Addresses context window limits in long conversations.

Lesson 2 • Building Consistent AI Personas

Defines persona attributes including name, expertise, tone, and behavioural guardrails in system prompts. Tests persona consistency under adversarial and off-topic user inputs.

Lesson 3 • Handling Edge Cases and Fallbacks

Prepares system prompts to gracefully handle out-of-scope requests, ambiguous inputs, and refusal scenarios. Designs fallback responses that maintain user trust and session continuity.

Lesson 1 • Privacy and Data Handling in Prompts

Covers risks of embedding personal, confidential, or proprietary data in prompts sent to external models. Teaches anonymisation and data minimisation techniques for safe prompt construction.

Lesson 2 • Evaluating Output Quality and Truthfulness

Provides a structured framework for assessing factual accuracy, completeness, and coherence in AI outputs. Builds a verification habit that complements prompt-level quality controls.

Lesson 3 • Responsible Use and Content Boundaries

Establishes content policy principles for professional AI use including harm avoidance and accuracy obligations. Connects organisational policy to practical prompt-level controls.

Lesson 4 • Understanding Prompt Injection and Jailbreaks

Defines prompt injection attacks and jailbreak techniques that manipulate model behaviour. Teaches defensive prompt design to resist adversarial inputs in deployed systems.

Lesson 1 • Retrieval-Augmented Generation Prompting

Integrates external retrieved documents into prompts to ground responses in current, verified information. Covers retrieval context formatting and citation instruction patterns.

Lesson 2 • Prompt Chaining and Pipelines

Connects sequential prompts where each output feeds the next stage of a multi-step workflow. Designs error-handling and validation checkpoints between pipeline stages.

Lesson 3 • Meta-Prompting and Self-Improvement

Uses prompts that instruct the model to generate, evaluate, or improve other prompts. Applies meta-prompting to automate prompt optimisation at scale.

Lesson 4 • Prompt Performance Evaluation at Scale

Introduces automated evaluation metrics, benchmark datasets, and A/B testing frameworks for large-scale prompt assessment. Connects evaluation results to iterative prompt improvement cycles.