Natural Language search

Middleware's AI-powered search feature is designed to simplify how customers interact with their data. Users can use natural language to search for data across dashboards.

How It Works

The Natural Language Search feature allows users to input queries in plain English. The system then interprets these queries and converts them into the appropriate query language to fetch the required data. However, it's essential to remember that context is key. The more detailed and specific the question, the more accurate the result.

Example of Effective Questions

• Good Example: "Show me hosts where memory usage is less than 10% and OS type is Linux."
• Bad Example: "I need Linux hosts with low memory usage."

When forming queries, be sure to include key details like relevant metrics, conditions, and attributes, and avoid ambiguity or vague terms.

Sample Queries

• "Show me hosts where CPU utilization is less than 10%."
• "Show me CPU utilization of the Docker container with the container name 'middleware'."
• "What is the status of the host filesystem?"
• "Show me the host with the highest memory usage."
• "Has the k8s container restarted in last 24 hours?"

How to Use

1. Navigate to the Infrastructure dashboard

2. Click on the search bar and enter your question

Enter your question and click on play button or hit enter

3. The results will be displayed as follows

Data Privacy and Security

It's important to note that our natural language search feature operates without accessing or sharing any customer data. This means:

• No customer data is ever shared with any Large Language Models (LLMs) or AI systems.
• The feature processes queries without needing to access your actual data.
• Your sensitive information remains secure and private at all times.

This design ensures that you can leverage the power of natural language queries while maintaining the highest standards of data protection and confidentiality.

Important Note

Please note that the Natural Language Search feature is currently in beta. While it strives for accuracy, it may occasionally produce errors or unexpected results. We encourage users to provide feedback to help us improve this feature.