Skip to main content
Version: 14.x

Leveraging the Software Catalog with Mia-Assistant

The Scenario​

An organization has fully embraced Mia-Platform, managing dozens of projects, hundreds of microservices, and a complex cloud infrastructure. As the platform scales, developers and operators find it increasingly difficult to find information and troubleshoot issues efficiently.

The Challenge​

  • Information Overload: Finding specific information—like the owner of a service, its current version in production, or its resource consumption—requires navigating through multiple pages in the Console or querying different systems.
  • High Cognitive Load for Troubleshooting: When an issue occurs, an operator needs to manually correlate information from different sources. For example, to debug a failing pod, they might need to check its logs, its deployment configuration, recent commits in its Git repository, and the status of its dependencies.
  • Onboarding Complexity: New team members are overwhelmed by the amount of information and struggle to understand the architecture and dependencies of the systems they are working on.
  • Need for Actionable Insights: The platform generates a vast amount of data, but turning that data into actionable insights (e.g., identifying optimization opportunities or potential risks) is a manual and time-consuming process.

The Solution with Mia-Platform​

The company enables the AI features of Mia-Platform Console, turning its platform into an intelligent system that can be queried and managed through natural language.

  1. Enriching the Software Catalog with Real-Time Data: The journey begins with data. The team configures the Integration Connector Agent to automatically scrape data from their cloud providers (like Azure) and DevOps tools (like GitLab). This data is used to enrich the Software Catalog, turning it from a static list of components into a dynamic, real-time representation of their entire software ecosystem. The catalog now knows about every deployed resource, its runtime status, its configuration, and its relationships with other components.

  2. Enabling AI Capabilities: In the Company Settings, a Company Owner enables the AI Settings. This grants Mia-Assistant, the AI-powered assistant within the Console, secure access to the enriched data from the Software Catalog and the project configurations.

  3. Interacting with the Platform via Mia-Assistant: Now, developers and operators can use Mia-Assistant to get immediate answers and perform actions:

    • A developer preparing for a release can ask: "@mia show me the DORA metrics for the checkout project in the last 30 days." Mia-Assistant queries the platform data and provides an instant report on deployment frequency and change failure rate.
    • An operator investigating an alert can use the /debug command: "/debug the payment service in production, are there any errors in the logs?" Mia-Assistant fetches the latest logs from the payment-service pod, analyzes them for error patterns, and provides a summary of its findings, often suggesting a root cause.
    • A new team member trying to understand the architecture can ask: "@mia what services does the order-service depend on?" The assistant uses the relationship data in the catalog to provide a clear dependency graph.
    • A platform engineer can even perform actions: "@mia deploy the latest version of the user-profile service to the staging environment." After a confirmation prompt, Mia-Assistant can trigger the deployment pipeline.

The Outcome​

  • Drastically Reduced Time to Information: Team members no longer waste time hunting for information. They get immediate, context-aware answers to their questions through a simple, conversational interface.
  • Accelerated Troubleshooting: The /debug command has become the first step in any troubleshooting process. It automates the initial data gathering and analysis, allowing operators to identify the root cause of issues much faster.
  • Democratized Platform Knowledge: The institutional knowledge about the platform is no longer confined to a few senior engineers. Anyone can now query the system and get the information they need, which has significantly improved onboarding and collaboration.
  • From Data to Actionable Intelligence: The platform is no longer just a system for managing configurations; it's an intelligent partner that helps teams make better decisions. Mia-Assistant turns raw data into actionable insights, proactively highlighting risks and suggesting improvements.

By integrating AI at its core, Mia-Platform Console transforms the developer and operator experience, moving from a traditional UI-based interaction to a conversational, intelligent, and highly efficient way of managing the entire software lifecycle.