Prompt Engineering Course

Lesson 1 • Types of Language Model Tasks

Categorizes tasks by cognitive demand: classification, extraction, generation, transformation, and reasoning. Helps learners match prompt strategy to task type.

Lesson 2 • Reading and Evaluating Model Output

Teaches systematic output inspection: identifying hallucinations, incomplete answers, and format drift. Grounds evaluation skills used throughout the entire course.

Lesson 3 • How Large Language Models Work

Covers tokenization, next-token prediction, and probability distributions in plain terms. Establishes the mental model needed to reason about why prompts succeed or fail.

Lesson 1 • Instruction Formatting Strategies

Covers markdown, delimiters, numbered lists, and XML-style tags to structure prompts. Shows how formatting reduces ambiguity and improves output consistency.

Lesson 2 • Few-Shot Prompting with Examples

Demonstrates how labeled input-output examples steer model behavior and format. Teaches example selection, ordering, and quantity trade-offs.

Lesson 3 • Negative and Constraint Prompting

Uses explicit exclusions and boundary conditions to prevent unwanted outputs. Complements positive instructions by narrowing the model's response space.

Lesson 4 • Role and Persona Prompting

Assigns expert identities to the model to shift tone, vocabulary, and reasoning style. Connects persona design to audience and task requirements.

Lesson 1 • Chain-of-Thought Fundamentals

Introduces the mechanism by which intermediate reasoning steps improve final answer quality. Establishes why explicit reasoning outperforms direct-answer prompts on hard tasks.

Lesson 2 • Decomposition and Step-Back Prompting

Breaks complex problems into sub-questions and uses abstraction to improve reasoning. Extends chain-of-thought to tasks requiring planning and hierarchical thinking.

Lesson 3 • Self-Consistency and Voting

Generates multiple reasoning paths and selects the most consistent answer. Introduces sampling strategies that reduce variance on high-stakes outputs.

Lesson 1 • Structured Data Output Prompting

Instructs models to return JSON, CSV, or key-value pairs with consistent schema. Covers schema definition, required fields, and validation strategies.

Lesson 2 • Long-Form Document Prompting

Structures prompts for reports, proposals, and articles with defined sections. Addresses length control, heading hierarchy, and tone consistency across long outputs.

Lesson 3 • Conditional and Dynamic Output Logic

Embeds if-then logic and branching instructions within prompts. Enables adaptive outputs that respond to varying input conditions without code.

Lesson 4 • Code Generation Prompting

Guides models to write functional, readable code with correct language syntax. Covers language specification, docstring requirements, and error-handling instructions.

Lesson 1 • Building a Prompt Evaluation Rubric

Creates task-specific scoring criteria for accuracy, format, tone, and completeness. Enables objective comparison across prompt versions and team members.

Lesson 2 • Prompt Versioning and Documentation

Establishes naming conventions, changelogs, and storage practices for prompt libraries. Prevents regression and enables team-wide reuse of proven prompts.

Lesson 3 • Diagnosing Prompt Failures

Classifies failure modes: ambiguity, missing context, format errors, and hallucination. Provides a diagnostic checklist that guides targeted revision.

Lesson 1 • Meta-Prompting and Prompt Generation

Uses the model to generate, critique, and optimize prompts automatically. Accelerates prompt development and surfaces non-obvious phrasings.

Lesson 2 • Retrieval-Augmented Prompting

Injects retrieved documents or facts into prompts to ground responses in verified sources. Reduces hallucination on knowledge-intensive tasks.

Lesson 3 • Self-Critique and Reflection Prompts

Instructs the model to evaluate and revise its own output within a single session. Produces higher-quality final answers without external feedback loops.

Lesson 4 • Prompt Chaining and Pipelines

Connects sequential prompts where each output feeds the next stage. Covers handoff design, error propagation, and pipeline debugging.

Lesson 1 • Customer-Facing Conversational Prompts

Builds system prompts for chatbots handling support, sales, and FAQ scenarios. Covers tone guardrails, escalation logic, and brand voice consistency.

Lesson 2 • Data Analysis and Insight Extraction

Prompts models to interpret tabular data, identify trends, and generate narrative summaries. Bridges prompt engineering with analytical reporting needs.

Lesson 3 • Content Creation and Editing Workflows

Designs prompt sequences for drafting, editing, summarizing, and repurposing content. Connects prompt design to editorial quality standards.

Lesson 4 • Automation and Workflow Integration

Embeds prompts into automated pipelines, APIs, and no-code tools. Covers parameterization, dynamic variable injection, and output routing.

Lesson 1 • Bias, Fairness, and Harm Reduction

Identifies how prompts can amplify or mitigate model bias and produce harmful outputs. Teaches proactive design choices that reduce discriminatory or unsafe responses.

Lesson 2 • Prompt Injection and Security

Explains adversarial prompt attacks that hijack model behavior in deployed systems. Teaches defensive prompt design and input sanitization strategies.

Lesson 3 • Building an Organizational Prompt Strategy

Designs team-level standards for prompt ownership, review, and continuous improvement. Connects individual prompting skill to scalable organizational capability.

Lesson 4 • Privacy and Data Handling in Prompts

Covers risks of embedding personal or confidential data in prompts sent to external models. Establishes data minimization and anonymization practices for prompt design.