Optimization Token Setup
Enable self-improving AI and CI/CD deployment by selecting an optimization token. This single configuration activates all optimization features automatically.
Once you select an optimization token, self-improving AI and CI/CD deployment activate immediately. The AI will start analyzing your evaluation data and generating optimized prompts automatically.
How It Works
Setting up optimization is simple:
Go to Settings
Navigate to Settings → Model Tokens in your dashboard
Select Optimization Token
Choose which AI model token to use for optimization analysis
Automatic Activation
Self-improving AI and CI/CD deployment activate immediately
Start Optimizing
The system begins analyzing evaluation data and generating improvements
Token Selection Process
Simple Setup
Navigate to Model Tokens:
-
Go to Settings
- Click Settings in your dashboard
- Select Model Tokens
- Find the Optimization section
-
Select Your Token
- Choose from your existing model tokens
- Click Use for Optimization
- Confirm your selection
What You’ll See:
- Your existing evaluation tokens (can be reused for optimization)
- Any additional model tokens you’ve configured
- Clear recommendations for best optimization performance
- Simple one-click selection process
Recommended Tokens
Best Performance Options
What Activates Automatically
When you select an optimization token, these features activate instantly:
Self-Improving AI:
- Analyzes your evaluation data for quality issues
- Generates improved prompt versions automatically
- Creates A/B testing configurations
Background A/B Testing:
- Tests new optimizations using cost-efficient sampling
- Provides statistical validation and performance comparisons
Release Hub:
- Shows optimization results and enables prompt comparison
- Provides deployment recommendations and tracks version history
CI/CD Integration:
- Makes optimized prompts available via SDK
- Enables instant deployment with rollback capabilities
Using Your Optimizations
Once optimization is active, you can access results through three main areas:
A/B Testing Results:
- Performance comparisons with statistical confidence
- Improvement metrics and testing progress
Release Hub:
- Compare current vs optimized prompts
- Deploy optimizations and track history
SDK Integration:
- Fetch optimized prompts in your applications
- Automatic updates with seamless integration
Next Steps
Ready to start optimizing your prompts?
- Go to Settings → Model Tokens to select your optimization token
- Monitor A/B testing results as they become available
- Use Release Hub to deploy optimizations
- Integrate with SDK to fetch optimized prompts
One-click activation! Select your optimization token and all optimization features activate immediately. Self-improving AI will start analyzing your data and generating improvements automatically.