Certified Prompt Engineering — Techniques
Chain-of-thought
Module: Reasoning Methods
Asking the model to reason step by step before answering boosts accuracy on logic and math.
In practice: 'Show your working, then state ANSWER: on a new line.'
Try It Yourself
'Show your working, then state ANSWER: on a new line.'
Lessons
▶️ 1. Chain-of-thought
▶️ 2. Tree-of-thought
▶️ 3. Self-consistency
▶️ 4. Decomposition
▶️ 5. Least-to-most prompting
▶️ 6. Reflection and self-critique
▶️ 7. Verification prompts
▶️ 8. Reasoning vs speed trade-off
▶️ 9. Designing a persona
▶️ 10. Expertise framing
▶️ 11. Tone and voice control
▶️ 12. Audience targeting
▶️ 13. Persona boundaries
▶️ 14. Multiple personas
▶️ 15. Consistency across a chat
▶️ 16. Persona for brand voice
▶️ 17. What is context engineering
▶️ 18. Managing the context window
▶️ 19. Grounding with sources
▶️ 20. Structured inputs
▶️ 21. Retrieval basics
▶️ 22. Memory strategies
▶️ 23. Avoiding context overload
▶️ 24. Context for long documents
▶️ 25. Requesting JSON
▶️ 26. Tables and markdown
▶️ 27. Controlling length
▶️ 28. Tone and style locks
▶️ 29. Enforcing schemas
▶️ 30. Step-limited output
▶️ 31. Multi-format output
▶️ 32. Stop sequences and markers
▶️ 33. Prompt chaining
▶️ 34. Sequential workflows
▶️ 35. Map-reduce prompting
▶️ 36. Conditional logic
▶️ 37. Branching prompts
▶️ 38. State passing
▶️ 39. Error handling in chains
▶️ 40. When to chain vs single prompt
Course Home