DevCoach 196: Prompt Engineering: Keterampilan Krusial Developer di Era AI

Halo, Coders! 👋

Di era di mana AI makin terintegrasi dalam proses software development, kemampuan memahami dan mengoptimalkan prompt menjadi nilai tambah yang signifikan bagi para developer.

Nah, melalui sesi Dev Coach kali ini, kita akan bahas bareng apa itu prompt engineering, kenapa keterampilan ini penting, serta bagaimana mengaplikasikannya secara efektif dalam aktivitas developer. Yuk, ikutan!

Transcript

1. Prompt Engineering: Keterampilan Krusial Developer di Era AI Nur Arifin Akbar Doctoral Researcher in AI Prompt Engineering
2. What We'll Cover Today Why Developers Need to Know Prompt Engineering What is Prompt Engineering? Best Practices for Developer Activities Q&A Session
3. Quick Poll: Your AI Experience How often do you currently use AI tools in your development work? • Never tried them • Occasionally (once a week or less) • Regularly (several times a week) • Daily
4. Why Developers Need to Know Prompt Engineering Visual: Developer working alongside AI assistant icon
5. The Development Landscape is Changing Statistics: • 70% of developers already use AI tools (2025) • Average time saved: 2-3 hours per day • Most common uses: Code completion, debugging, documentation
6. What's In It For You? Save Time Automate repetitive coding tasks Learn Faster Get explanations and examples instantly Write Better Code Get suggestions for optimization Reduce Bugs AI can spot patterns you might miss Stay Current Keep up with best practices
7. Future-Proof Your Career 1 Foundation for AI-augmented development 2 Required by forward-thinking companies 3 Differentiator in job market 4 Emerging skill in high demand Key Points: • Emerging skill in high demand • Differentiator in job market • Required by forward-thinking companies • Foundation for AI-augmented development "AI won't replace developers, but developers using AI will replace those who don't."
8. "But What About..." ❌ "AI will replace developers" ✅ AI amplifies developer capabilities ❌ "Generated code is low quality" ✅ Quality depends on prompt quality ❌ "It's too complex to learn" ✅ Basic skills can be learned in hours ❌ "It's just hype" ✅ Major companies investing billions
9. Quiz #1  DevCoach 196 Next >>>
10. Quiz #1  DevCoach 196 Manakah pernyataan benar dari statistik di samping? a. 70% kode yang dihasilkan AI memiliki kualitas yang buruk b. 70% developer setuju AI dapat menghemat 23 jam per hari c. 70% developer sudah menggunakan AI Tools Prompt Engineering Format jawaban: #quiz1-username-jawaban
11. What is Prompt Engineering? Human Prompt AI Output
12. Prompt Engineering in Plain English "Prompt engineering is the skill of writing clear instructions that help AI understand exactly what you want it to do." Analogy: Like giving directions to a very smart but very literal assistant
13. Anatomy of a Good Prompt Context What background information is needed? 1 Task What do you want the AI to do? 2 Requirements What constraints or specifications? 3 Format How should the output look? 4 Simple Example: "You are helping a Python developer. Create a function that sorts a list of dictionaries by a specific key. Use type hints and include a docstring."
14. The Difference Quality Makes ❌ Poor Prompt: "Write code for login" ✅ Good Prompt: "Create a Python Flask route for user login that validates email/password, uses bcrypt for password checking, returns JWT token on success, and handles errors gracefully" Result: Generic code vs. Production-ready implementation
15. Different Prompts for Different Tasks Instructional: "Create a function that..." Debugging: "Why does this code produce this error..." Explanatory: "Explain how this algorithm works..." Refactoring: "Improve this code by..." Review: "Check this code for potential issues..."
16. How AI Interprets Your Prompts AI responds to patterns in your language Specific words trigger specific behaviors Context influences interpretation Examples guide output format Tip: Think of AI as pattern matching, not true understanding
17. Your First Prompt is Just the Beginning Write initial prompt Analyze the output Identify what's missing or wrong Refine your prompt Repeat until satisfied 1 2 3 4 5 Example: Improving a function from basic to production-ready through iteration
18. Quiz #2  DevCoach 196 Next >>>
19. Quiz #2  DevCoach 196 Berikut adalah urutan dalam merancang prompt. Apa poin nomor 5? a. Analisis output yang dihasilkan model b. Refine prompt awal c. Ulangi proses refine prompt Prompt Engineering Format jawaban: #quiz2-username-jawaban 1 2 3 4 5
20. Best Practices for Developer Activities Visual: Code editor with AI suggestions panel
21. Writing Code with AI 1 Specify language and version: "Python 3.9" 2 Name frameworks/libraries: "using FastAPI" 3 Describe the purpose clearly 4 Include error handling requirements 5 Request comments for complex logic Live Example: Generate a REST API endpoint
22. AI as Your Debugging Partner Effective Debugging Prompts Include: • The problematic code • The exact error message • What you expected to happen • What actually happened • What you've already tried Template: "I have this [language] code: [code]. It produces this error: [error]. I expected [expectation] but got [actual result]. I've tried [attempts]. What's wrong?"
23. Comprehensive Test Generation Prompt Strategy: "Create unit tests for this function: [function code] Use [test framework] Include: - Happy path tests - Edge cases - Error scenarios - Performance tests if applicable" Result: Better coverage in less time
24. Documentation That Developers Will Actually Read Documentation Prompt Framework: • Purpose and overview • Parameters and return values • Usage examples • Common errors and solutions • Performance considerations Before: "Document this function" After: "Create developer documentation for this function including purpose, parameters with types, return value, 2-3 usage examples, and common gotchas"
25. AI as Your First Code Reviewer Review Checklist Prompts: • "Review this code for security vulnerabilities" • "Check for performance bottlenecks" • "Identify code style violations" • "Suggest readability improvements" • "Find potential bugs" Important: AI supplements, doesn't replace, human review
26. AI as Your Learning Companion Learning Prompts: • "Explain this code snippet in simple terms" • "What's the difference between X and Y?" • "Show me three ways to implement..." • "What are the pros and cons of..." • "Convert this code from X to Y language"
27. Learn From Others' Mistakes 1 Vague prompts → Be specific and detailed 2 No context → Always provide relevant background 3 Ignoring errors → Review all generated code 4 One-shot attempts → Iterate and refine 5 Blind trust → Always verify AI output
28. Let's Try It Together! Exercise: Write a prompt to generate a user authentication function Your Mission: • Include all four components (context, task, requirements, format) • Be specific about security needs • Specify your tech stack • Request error handling Share your prompts in the chat!
29. Continue Learning dicoding.com/academies/753
30. Remember These Points 1 Prompt engineering saves hours of development time 2 Quality prompts = quality outputs 3 Iteration is normal and expected 4 Specific techniques for specific tasks 5 Always verify AI-generated code Your Challenge: Use AI for one development task this week
31. Quiz #3  DevCoach 196 Next >>>
32. Quiz #3  DevCoach 196 Berikut ini yang merupakan common use case dari AI untuk software developer adalah … a. Menulis kode b. Reviewer kode c. Dokumentasi kode d. Semua jawaban benar Prompt Engineering Format jawaban: #quiz3-username-jawaban
33. Feedback! dicoding.id/devcoachfeedback Prompt Engineering
34. Thank You dicoding dicoding info@dicoding.com Prompt Engineering