Skip to main content

Get started in three steps

Start evaluating your AI agents with realistic user simulations in just a few minutes.

Step 1: Add Your Agent

Connect your AI agent to UserTrace for evaluation.
Add your voice agent connection:Telephony-based: Provide your voice agent’s phone number
App-based: Provide your WebRTC URL
Add your agent using either:Prompt-based: Paste your system prompt directly
API-based: Provide your Chat Completions API endpoint to your Agent Sandbox

Step 2: Create Scenarios

Describe your user in simple text and UserTrace automatically generates everything needed for testing.
Just describe your user in natural language:“30 year old female, recently moved from US to Germany, stressed about cultural differences at work”UserTrace will automatically create:Realistic personas with matching characteristics
Test scenarios relevant to this user type
Evaluation criteria tailored to their needs
All linked together for comprehensive testing

Step 3: Start Simulation

Simply select your agent and scenario - everything else is automatically configured.
  1. Select your agent from the dropdown
  2. Choose your scenario - linked personas and evals auto-populate
  3. Optionally add specific personas or evaluations you want to include
  4. Click “Start Simulation” and watch real-time results
  5. Get detailed feedback on functionality, safety, and user satisfaction

What You’ll Get

Within minutes of running your first simulation, you’ll receive:

Performance Insights

Detailed analytics on response quality, conversation flow, and user satisfaction metrics.

Issue Detection

Identify conversation breakdowns, safety concerns, and areas for improvement.

User Experience Analysis

Understand how real users would interact with your agent and where friction occurs.

Actionable Recommendations

Get specific suggestions for improving your agent’s responses and capabilities.

Next Steps

Ready to dive deeper? Explore our integration guides for Voice Agents or Custom Chat Agents.