Infrastructure tools usually give you a flat list of resources. Knowledge Tree gives you a graph.
Flat Lists Aren't Enough
A typical cloud inventory tells you: "You have 47 EC2 instances, 12 RDS databases, 3 VPCs." That's useful, but it doesn't tell you:
- Which EC2 instance connects to which RDS database
- Which VPC your Kubernetes cluster runs in
- What happens if you delete that security group
The Knowledge Graph
Knowledge Tree uses Apache AGE — a PostgreSQL extension that adds graph database capabilities. Every resource is a node. Every relationship is an edge.
This means you can ask questions like:
MATCH (ec2:EC2 {environment: 'production'})-[:CONNECTS_TO]->(rds:RDS)
RETURN ec2.name, rds.nameAnd get back meaningful answers about your infrastructure topology.
Why PostgreSQL?
We didn't want to add another database to your stack. By building on PostgreSQL with AGE and pgvector, you get:
- Graph queries (Cypher) for relationship traversal
- Vector search (pgvector) for semantic similarity
- Relational queries (SQL) for traditional filtering
- One database to manage, backup, and scale
What This Enables
The knowledge graph is what makes our auto-documentation actually useful. When we generate a Mermaid diagram of your infrastructure, it's not just a list — it shows the actual connections between resources.
Combined with LLM enrichment, you get natural-language explanations of your architecture that are always up-to-date.