Python · Beginner
AI & Prompt Engineering
Fifty-six lessons on prompts, hallucination, privacy, guardrails and the Python behind the academy’s own assistant, ending with a capstone you can show an employer.
Thinking about AI
- 1 What AI really is Concept —
- 2 Where AI helps an academy Concept —
- 3 Tokens and limits Concept —
- 4 Prompts are instructions Concept —
- 5 The four-part prompt Concept —
- 6 Give examples Concept —
- 7 Ask for a format Concept —
- 8 Hallucination Concept —
- 9 Never paste private data Concept —
- 10 Check the work Concept —
Prompt engineering
- 11 Python: your first AI call Python —
- 12 Sizing a request Python —
- 13 Temperature Python —
- 14 Max tokens and cost Python —
- 15 The system prompt Python —
- 16 Guardrails Python —
- 17 The FAQ bank Python —
- 18 Fallback answers Python —
- 19 Log and read Python —
- 20 The human in the loop Concept —
Build the assistant
- 21 Clean the text Python —
- 22 Keywords, not sentences Python —
- 23 Scoring a match Python —
- 24 Write the answer copy Python —
- 25 Link the answer Python —
- 26 The admission reply Python —
- 27 Build a fee table Python —
- 28 Answer an objection Python —
- 29 Follow up Python —
- 30 Measure quality Concept —
Content and quality
- 31 A working rule-based bot Python —
- 32 Add an answer without breaking anything Python —
- 33 Handle a missing answer Python —
- 34 When to add a real model Python —
- 35 Bias and fairness Concept —
- 36 Accessibility of chatbots Concept —
- 37 Privacy policy Concept —
- 38 Cost control Python —
- 39 Version your prompts Python —
- 40 Test with real questions Python —
Real work and portfolio
- 41 AI for notices Concept —
- 42 AI for lesson outlines Concept —
- 43 AI for practice questions Concept —
- 44 AI for student feedback Concept —
- 45 Where the job market is Concept —
- 46 Build a portfolio Concept —
- 47 Estimate tokens for a real prompt Python —
- 48 Read a number out of a message Python —
- 49 Summarise simply Python —
- 50 Schedule the reminder Python —
Capstone: the pieces of an agent
- 51 Estimate the cost of a prompt Python —
- 52 Fill a prompt template Python —
- 53 Score similarity with a dot product Python —
- 54 Chunk a document for retrieval Python —
- 55 Add a guardrail before the answer Python —
- 56 Plan an agent's steps Python —
Final exam and certificate
Twenty questions from the whole track. Pass mark is 60%. Retakes are free and the best score is the one printed on your certificate.
Create a free account to take the exam