Skip to content
NerdLevelTech
BeginnerPremium · 1,000 credits

The Prompt Engineering Path — Foundations

Learn the craft of prompting — from your first chat message to a tight, production-grade system prompt you can ship. Every example is a real captured model output, dated and attributed, including the ones where the model got it wrong. Day-to-day scenarios, no fluff. The gateway to the full Prompt Engineering learning path.

138 min42 lessons9 modulesAI & ML

What you'll learn

  • Write specific, testable prompts that consistently produce useful outputs
  • Apply the 5-slot prompt skeleton (role, task, context, input, output spec)
  • Use 10 core patterns: zero-shot, few-shot, chain-of-thought, role-play, format-lock, refusal-shape, repair, self-critique, decomposition, persona
  • Diagnose and repair bad outputs without restarting from scratch
  • Move from chat UI to API system prompts with confidence
  • Write tight 5-part system prompts for production use
  • Recognise hallucinations, knowledge cutoffs, and trigger 'I don't know' patterns

What's included

Video Lessons
—
Interactive Exercises
—
Charts & Infographics
Included
Quizzes
9
Certificate
Included

Course Content

  1. Why prompts matter

    5 lessons · 15 min
    • What an LLM Actually DoesNo signup3 min
    • Locked until you enrollVague vs Specific: Same Model, Different Result4 min
    • Locked until you enrollAn LLM Is Not a Search Engine3 min
    • Locked until you enrollThe Skill Stack2 min
    • Locked until you enrollYour First Deliberately Specific Prompt3 min
    • Quiz: Module 1: Why Prompts Matter

  2. The anatomy of a prompt

    5 lessons · 16 min
    • Locked until you enrollThe Five-Slot Prompt Skeleton3 min
    • Locked until you enrollThe Role Slot2 min
    • Locked until you enrollTask vs Context3 min
    • Locked until you enrollThe Output Spec — and the Format-Lock Trick4 min
    • Locked until you enrollWorked Example: All Five Slots Together4 min
    • Quiz: Module 2: Anatomy of a Prompt

  3. Your first 10 patterns

    6 lessons · 21 min
    • Locked until you enrollZero-Shot: Just Ask3 min
    • Locked until you enrollFew-Shot: Show, Don't Just Tell4 min
    • Locked until you enrollPersona: Same Question, Two Different Voices4 min
    • Locked until you enrollChain-of-Thought Light: Show Your Working3 min
    • Locked until you enrollSelf-Critique: Make the Model Grade Itself4 min
    • Locked until you enrollDecomposition: Break the Big Task into Small Steps3 min
    • Quiz: Module 3: Your First Patterns

  4. Day-to-day playbook (verified)

    5 lessons · 19 min
    • Locked until you enrollEmail Rewrite — From Vague to Shippable4 min
    • Locked until you enrollMeeting Summary — Vague Prompt vs Structured Prompt4 min
    • Locked until you enrollLinkedIn Rewrite — A Live Iteration4 min
    • Locked until you enrollCustomer Reply (No System Prompt)3 min
    • Locked until you enrollCustomer Reply (With a Tight System Prompt)4 min
    • Quiz: Module 4: Day-to-Day Playbook

  5. Diagnosing bad outputs

    4 lessons · 13 min
    • Locked until you enrollRewrite, Don't Restart3 min
    • Locked until you enrollThe Four Diagnostic Questions3 min
    • Locked until you enrollWorked Repair: The LinkedIn Rewrite, Fixed4 min
    • Locked until you enrollWhen to Start Clean3 min
    • Quiz: Module 5: Diagnosing Bad Outputs

  6. From chat to API: the system slot

    4 lessons · 13 min
    • Locked until you enrollWhat changes when you leave the chat tab3 min
    • Locked until you enrollThe system slot — your sticky instruction3 min
    • Locked until you enrollChat vs API: same task, two outputs4 min
    • Locked until you enrollTemperature — and why it was taken away3 min
    • Quiz: Module 6: From chat to API

  7. The tight system prompt

    5 lessons · 16 min
    • Locked until you enrollThe 5-part skeleton, scaled up3 min
    • Locked until you enrollRole and Capabilities — sharp, not generic3 min
    • Locked until you enrollConstraints — building the Bayt Coffee voice3 min
    • Locked until you enrollFormat and Examples — what good output looks like3 min
    • Locked until you enrollBayt Coffee — the full prompt, the live reply4 min
    • Quiz: Module 7: The tight system prompt

  8. Knowing what the model doesn't know

    4 lessons · 12 min
    • Locked until you enrollKnowledge cutoffs — what the model has not seen3 min
    • Locked until you enrollHedge language — built into the system prompt2 min
    • Locked until you enrollThe 'I don't see that in the document' trigger4 min
    • Locked until you enrollThe four hallucination patterns to test for3 min
    • Quiz: Module 8: Knowing what the model doesn't know

  9. Capstone — three production prompts

    4 lessons · 13 min
    • Locked until you enrollThe capstone — three production prompts of your own3 min
    • Locked until you enrollThe 10-point rubric3 min
    • Locked until you enrollThree example scenarios4 min
    • Locked until you enrollWhat good and bad submissions look like3 min
    • Quiz: Module 9: Capstone