Skip to main content

Exploring the Ontology

After discovery runs and entities are populated, you can explore your organization's knowledge graph through the interactive ontology visualization.

Prerequisites​

Before exploring the ontology, make sure you have:

  • Created an environment
  • Configured at least one molecule (GitHub, AWS, etc.)
  • Run discovery to populate entities

If discovery hasn't run yet, follow the First Environment guide.

Accessing the Ontology​

Navigate to your environment at app.sixdegree.ai. The ontology view is displayed by default on the main dashboard. You'll see your discovered entities as nodes and relationships as edges.

What You'll See​

The ontology visualization uses circles to represent entities like repositories, services, and users. Lines between nodes show relationships, and different colors indicate different entity types.

Basic Controls​

Use your mouse wheel or scroll gesture to zoom in and out of the graph. Click and drag the background to pan around and explore different areas. Click any node to see detailed information about that entity. If you need to reset your view, use the "Fit to View" button to center and zoom to show all entities.

Filtering​

You can reduce clutter by filtering what's displayed in the graph. Toggle checkboxes to show or hide specific entity types. For example, hide users to focus only on repositories and services. Use the search box to find specific entities by name - the graph will highlight or center on matching entities, making it easy to quickly locate a specific service or repository.

Understanding Entity Details​

When you click on any node, you'll see detailed information including the entity's name, type (Repository, Service, User, etc.), which molecule discovered it, and when it was first found and last updated.

The details panel also shows all properties specific to that entity, such as URLs, endpoints, configuration details, and metadata from the source system. You'll see both outgoing relationships (connections from this entity to others) and incoming relationships (connections from other entities to this one).

Common Patterns to Look For​

As you explore the ontology, you'll notice patterns emerging. Ownership structures show organizations owning multiple repositories, teams managing services, and users belonging to groups. Deployment chains reveal how repositories are built into Docker images, which are then deployed by Kubernetes or Argo CD into clusters.

Dependency networks illustrate how services depend on other services, shared libraries used across repositories, and infrastructure dependencies. Team collaboration patterns show users contributing to repositories, shared ownership of services, and cross-team dependencies.

Exploring by Scenario​

Finding What a Team Owns​

Search for a team or organization name and click the node to select it. Look at the outgoing relationships to see all the entities that team owns. Pay attention to OWNS or BELONGS_TO relationships to understand the full scope of their resources.

Understanding Service Dependencies​

Search for your service name and click to select it. Examine the relationships to see which services this one depends on (DEPENDS_ON) and which services depend on it (incoming DEPENDS_ON relationships). Follow the dependency chain to understand the full impact of changes.

Discovering Infrastructure​

Filter the view to show only infrastructure types like clusters and instances. Look for deployment relationships that connect code repositories through build processes to runtime environments. This helps you trace the complete path from source code to production.

Identifying Orphaned Resources​

Look for isolated nodes with few or no connections. These might represent unused services or forgotten resources that could be cleaned up or better documented.

Refreshing Data​

Discovery runs automatically on schedule based on how you've configured each molecule. Entities and relationships update automatically as they're discovered. If you need immediate updates, you can trigger discovery manually from Environment Settings and check the discovery status to see when it last ran.

Next Steps​

Now that you understand the ontology:

Tips for Effective Exploration​

Start small by beginning with one or two molecule types. Understand that data before adding more, and gradually expand your discovery scope. Use filters strategically by hiding entity types you don't need right now, allowing you to focus on one subsystem at a time and create mental models of different areas.

Follow the relationships rather than just looking at individual entities. Trace connections to understand your architecture, and identify critical paths and dependencies. Document your findings by using chat to ask questions about what you discover, sharing interesting patterns with your team, and updating documentation based on your discoveries.

Troubleshooting​

No Entities Visible​

Check that discovery has completed in Environment Settings, filters aren't hiding everything, molecules are configured correctly, and API credentials are valid.

Graph Won't Load​

Try refreshing the page, clearing your browser cache, using filters to reduce entity count, or checking the browser console for errors.

Can't Find Specific Entity​

Verify that the entity actually exists in the source system, the molecule discovery includes that entity type, your search query is correct, and the entity type filter isn't hiding it.

Advanced Exploration​

Look for clusters of highly connected entities to understand patterns in your infrastructure. Identify hub nodes with many relationships - these are often critical components. Find isolated or weakly connected components that might need attention.

For impact analysis, select a critical entity and explore all its relationships to understand the blast radius of potential changes and identify downstream dependencies.

When mapping your architecture, start from user-facing services and follow deployment relationships backward. Trace back to source repositories to map the complete deployment pipeline from code to production.

Need Help?​