Stage 9 of 12 · Advanced

Advanced Prompting

The full prompting process and every technique worth knowing.

Go past good prompts to a repeatable process: define the goal, gather context, draft, test, iterate, evaluate and ship. Then master the techniques experts use every day, from few-shot examples and schemas to chaining, reasoning models, self-consistency, meta-prompting, evals, multimodal inputs and safety.

0/11

modules

106

minutes total

Modules

1
The full prompting process
10 min
Seven steps that take any task from a vague idea to a prompt you can trust and reuse.
Start here →
2
Few-shot prompting and examples
9 min
How examples teach format, style and judgement, how to choose them, and when to leave them out.
3
Roles and system prompts
9 min
What system prompts and personas really do, and how to write one that holds up across a whole conversation.
4
Structured output: JSON and schemas
10 min
Get machine-readable output you can rely on, design schemas that help the model, and validate everything.
5
Prompt chaining and decomposition
10 min
Split hard tasks into steps with clean hand-offs, and know when one prompt is enough.
6
Chain-of-thought and reasoning models
10 min
When step-by-step reasoning helps, when it doesn't, and how to prompt models that already think before they answer.
7
Self-consistency and multiple samples
8 min
Ask more than once: majority voting for answers, best-of-n for open-ended work, and disagreement as a warning light.
8
Meta-prompting and prompt templates
9 min
Have models write and improve your prompts, and turn the winners into reusable templates with variables.
9
Testing prompts: golden sets, rubrics and judges
11 min
Replace vibes with evidence: build a golden set, write rubrics, and use model judges without being fooled by them.
10
Prompting with images, audio and video
9 min
Get accurate answers from screenshots, photos, charts, recordings and clips, and know where multimodal models still slip.
11
Safety: prompt injection, hallucinations and verification
11 min
Keep untrusted content from steering your AI, catch made-up facts, protect private data and verify before you act.