🧪 Step 3: Create and Run an Experiment

Experiments test your AI for vulnerabilities and behavioral issues. Let's create your first test.

3.1 Start Experiment Creation

1. Go to your project dashboard
2. Click "Create Experiment"
3. Choose your experiment type:
   • Adversarial Testing: Tests for security vulnerabilities
   • Behavioral QA: Tests for functional correctness

3.2 Configure Your Experiment

Basic Settings:

• Name: Choose a descriptive name (e.g., "Security Test - Customer Support")
• Description: Explain what you're testing
• Model Provider: Select the provider you set up in Step 1
• Language: Choose the primary language for testing

Advanced Settings:

• Conversation Turn:
  • Single-Turn: Test individual prompts
  • Multi-Turn: Test full conversations
• Testing Level:
  • Quick: ~500 tests, fast feedback
  • Thorough: ~1200 tests, balanced coverage
  • Comprehensive: ~2000+ tests, deep analysis

Integration Configuration:

Configure how ai+me connects to your AI:

1. Thread Initialization (if applicable):

   • Endpoint URL for starting conversations
   • Headers and payload format

2. Chat Completion:

   • Endpoint URL for sending messages
   • Headers and payload format
   • Enable/disable streaming

3. Streaming Configuration:

   • Enable Streaming: If your AI returns responses in real-time chunks
   • Disable Streaming: If your AI returns complete responses at once

3.3 Launch Your Experiment

Click "Create Experiment" to start testing. Your experiment will:

1. Initialize: Set up the testing environment
2. Generate Tests: Create contextual test scenarios
3. Execute Tests: Run prompts against your AI
4. Analyze Results: Evaluate responses for issues
5. Generate Report: Compile findings and insights

⏱️ Expected Duration: 5-30 minutes depending on testing level and model response time.