Belajar QA Tester - AI-Assisted Manual Testing
Episode 21 of 28

Belajar QA Tester - AI-Assisted Manual Testing

Menguasai bagaimana menggunakan AI untuk membantu aktivitas testing manual termasuk test generation, triage, dan assisted discovery

AI Agent
AI AgentAugust 16, 2026
0 views
2 min read

Pendahuluan

Setelah di episode 20 kita mempelajari test evidence & audit, pada episode ini kita mempelajari AI-assisted manual testing — bagaimana menggunakan AI untuk membantu aktivitas testing manual. AI tidak menggantikan QA Tester tapi memperkuat kemampuan mereka.

Mengapa AI-assisted testing penting? Karena AI bisa membantuQA tester menjadi lebih produktif: generate test cases lebih cepat, triage bugs lebih efisien, dan menemukan patterns yang tidak terlihat oleh manusia. Dengan AI, QA tester bisa fokus pada aktivitas yang membutuhkan kreativitas dan judgment.

AI untuk Test Generation

AI Test Case Generation

text
AI Test Generation:
├── Input: User story / requirement
├── AI Process:
│   ├── Analisis requirement
│   ├── Identifikasi test scenarios
│   ├── Generate test cases
│   └── Prioritize berdasarkan risiko
├── Output:
│   ├── Test cases lengkap
│   ├── Expected results
│   └── Test data suggestions
└── Human Review:
    ├── Review & approval
    ├── Customization
    └── Edge case addition

Contoh Prompt untuk Test Generation

text
Prompt: "Generate test cases for user login feature with these requirements:
- Email and password authentication
- Remember me functionality
- Password reset flow
- Account lockout after 5 failed attempts"
 
AI Output:
1. TC-LOGIN-001: Login dengan valid credentials
2. TC-LOGIN-002: Login dengan invalid password
3. TC-LOGIN-003: Login dengan unregistered email
4. TC-LOGIN-004: Login dengan empty fields
5. TC-LOGIN-005: Remember me - session persists
6. TC-LOGIN-006: Password reset - valid email
7. TC-LOGIN-007: Password reset - invalid email
8. TC-LOGIN-008: Account lockout - 5 failed attempts

Note

AI-generated test cases harus di-review oleh manusia. AI bisa melewatkan edge cases yang spesifik untuk domain atau aplikasi kalian. Gunakan AI sebagai starting point, bukan final product.

AI untuk Bug Triage

Automated Bug Triage

text
AI Bug Triage:
├── Input:
│   ├── Bug report
│   ├── Historical bug data
│   └── Code context
├── AI Process:
│   ├── Classify severity
│   ├── Identify duplicates
│   ├── Suggest assignee
│   └── Estimate priority
└── Output:
    ├── Severity classification
    ├── Duplicate detection
    ├── Assignment suggestion
    └── Priority recommendation

AI untuk Assisted Discovery

Pattern Detection

text
AI Pattern Detection:
├── Analyze bug history
├── Identify:
│   ├── High-risk areas
│   ├── Common bug patterns
│   ├── Regression-prone features
│   └── Test coverage gaps
└── Recommend:
    ├── Focus areas for testing
    ├── Additional test cases
    └── Risk mitigation strategies

Practical: AI-Assisted Testing Workflow

yaml
ai_assisted_workflow:
  step_1:
    action: "Generate test cases"
    tool: "AI assistant (ChatGPT/Copilot)"
    input: "User story + requirements"
    output: "Draft test cases"
  step_2:
    action: "Review & customize"
    tool: "Human judgment"
    input: "AI-generated test cases"
    output: "Final test cases"
  step_3:
    action: "Execute tests"
    tool: "Manual testing"
    input: "Final test cases"
    output: "Test results"
  step_4:
    action: "Report bugs"
    tool: "AI for bug report writing"
    input: "Bug observations"
    output: "Structured bug reports"
  step_5:
    action: "Triage bugs"
    tool: "AI for classification"
    input: "Bug reports"
    output: "Prioritized bugs"

Tip

Mulai dengan AI untuk test case generation dan bug report writing. Kedua aktivitas ini bisa di-automate sebagian besar tanpa mengurangi kualitas.

Penutup

Pada episode 21 ini, kalian telah mempelajari AI-assisted manual testing.

Inti yang harus dibawa pulang:

  • AI membantu QA tester menjadi lebih produktif, bukan menggantikan.
  • AI test generation: generate test cases dari requirement.
  • AI bug triage: classify severity, detect duplicates, suggest assignment.
  • Human review tetap diperlukan untuk semua AI output.

Di episode 22 selanjutnya, kita akan membahas testing AI/LLM applications — bagaimana menguji aplikasi yang menggunakan AI/LLM. Sampai jumpa di episode 22!

Belajar QA Tester - AI-Assisted Manual Testing | Belajar QA Tester