6. Next Steps

You have installed GQLDB, loaded a graph, queried and traversed it, run an algorithm, and driven it all from an AI agent. Here is where to go deeper on each thread.

Go Deeper on the Query Language

• ISO GQL: the full language reference: pattern matching, data manipulation, functions, expressions, and transactions.
• Stored Procedures: package reusable logic you can call from your queries.

Model Your Data Properly

• Open Graphs: the schema-free default used in this guide, where labels and properties are created on the fly.
• Closed Graphs: enforce a schema of node and edge types for production data integrity.
• Ontology: model RDF data with OWL semantics, inference, and validation.

Analyze and Learn

• Graph Algorithms: the full catalog across centrality, community detection, similarity, pathfinding, and embeddings.
• Computing Engine: the in-memory engine that accelerates algorithms and traversals on large graphs.
• Machine Learning: in-database pipelines for node classification and link prediction.
• AI & Vectors: vector search and AI-powered graph functions.

Build Applications

• Drivers: Java, Python, Go, and Node.js SDKs, each with its own quick start.
• Ultipa MCP: the complete AI-agent integration reference.

Run It in Production

• Operations: deployment topologies, clustering, monitoring, and performance tuning.
• Backup & Restore: durability and disaster recovery.
• Access Control: role-based access control and security.

Get Help

• Report a documentation issue or suggest a change on GitHub.
• For product support, contact [email protected].