Skip to main content

​
DataGyro Playground

The DataGyro Playground is part of our Search vertical and provides an interactive environment where you can test and refine your natural language queries directly in your browser. It’s designed to help you perfect your queries and preview exactly how they’ll perform in your LLM-powered applications before writing any code.

​
Accessing the Playground

To access the Playground:
  1. Log in to your DataGyro account at app.datagyro.com
  2. Navigate to the “Playground” tab in the main navigation

​
Using the Playground

The Playground offers a streamlined interface for testing your natural language retrieval queries:
  1. Select an API Key - Choose which API key to use for authentication
  2. Select a Collection - Choose the collection containing your processed data
  3. Enter Your Query - Type your natural language question or search phrase
  4. Adjust Parameters - Optionally modify parameters like result limit or model selection
  5. Click “Search” - Run your query and see results in real-time
DataGyro Playground Interface

​
Understanding Results

The Playground displays query results in the same JSON format as the API response, giving you an exact preview of what your application will receive: 
{
  "results": [
    {
      "content": "DataGyro supports PostgreSQL and MySQL databases as data sources, with more connectors coming soon.",
      "metadata": {
        "source": "documentation.pdf",
        "page": 12,
        "section": "Data Sources"
      },
      "score": 0.94
    },
    // Additional results...
  ],
  "total_count": 8,
  "execution_time_ms": 124
}
Each result includes:
  • Content: The relevant text chunk from your data source
  • Metadata: Source information and context (varies based on data source type)
  • Score: A relevance score between 0-1 indicating match quality
  • Performance metrics: Total result count and execution time
This immediate feedback helps you understand how well your query is performing and allows you to refine it for optimal retrieval.

​
Benefits of Using the Playground

The Playground serves as a critical development tool in your natural language retrieval workflow:
  • Rapid Iteration: Test and refine queries without modifying your application code
  • Immediate Feedback: See exactly how your queries perform with real-time results
  • Query Optimization: Experiment with different phrasings to find the most effective formulations
  • Parameter Testing: Try different limits, model sizes, and other parameters to balance performance and cost
  • Simplified Implementation: Generate production-ready code snippets in your preferred language
  • Data Exploration: Gain insights into how your data is structured and best accessed
  • Troubleshooting: Identify and resolve retrieval issues before they affect your production application

​
Example Workflows

Here are some common ways developers use the Playground:

​
Query Refinement Workflow

  1. Select your API key and collection
  2. Start with a basic query: “Show me product pricing information”
  3. Review the initial results
  4. Refine your query to be more specific: “What is the monthly cost of the enterprise plan for 50+ users?”
  5. Compare the improved results
  6. Further refine if needed: “What’s included in the enterprise plan at the 50+ user tier?”
  7. Once satisfied, copy the code snippet for your application