Multi-turn test runs take longer to complete since each scenario involves multiple conversation steps
Quickstart
Test multi-turn AI conversations
Evaluate AI chat interactions automatically using conversation simulation, without code changes
Simulate and test multi-turn conversations with your AI agent using Maxim Workflows. Instead of manual testing, our simulation engine interacts with your agent based on predefined configurations.
Configure your HTTP endpoint
Add your API endpoint in Workflows. Configure request headers and body parameters as needed.
Configure test settings
Create a test run with:
- A dataset containing test scenarios
- Relevant evaluators for chat quality
- Simulation settings for conversation flow
- Optional columns for additional test parameters
Review simulation results
Analyze the test report to understand conversation quality and performance metrics.
Test your AI application via an API endpoint
Run your first test on an AI application via HTTP endpoint with ease, no code changes needed.
Test your first Prompt Chain
Test your agentic workflows using Prompt Chains with Datasets and Evaluators in minutes. View results across your test cases to find areas where it works well or needs improvement.