Welcome to GraphForge!
GraphForge is Graphicient’s data engineering studio. It lets you design your ontology (the structure of your knowledge graph), document it for the AI assistant, then load your data into it from files or databases — no prior knowledge of graph databases required.
What you can do with GraphForge
- Design and edit your ontology visually: objects, relations, groups, labels and interfaces
- Import and export an ontology (JSON, OWL + SKOS)
- Enrich the ontology for the AI assistant: business glossary, quality score
- Configure the environment’s AI agent: identity, instructions, starter questions, guardrails, skills
- Connect your data: Excel/CSV files or databases (Snowflake, Databricks, PostgreSQL)
- Transform data in Python before loading it
- Map sources to the ontology, run or schedule the load, and trace where data comes from
Getting started
GraphForge is organized into 4 modes, available from the bar at the top of the screen:
- Ontology — design the structure of your graph
- Semantic — consult the reference knowledge: overview, glossary, quality
- AI Agent — configure the AI assistant’s behavior
- Data — connect, transform and load your real data
You can switch freely from one mode to another, without reloading.
A typical journey
- I model my domain in Ontology (objects, relations, first groups)
- I check the quality score to see what is missing
- I enrich the least well-described elements (description, synonyms, business rules)
- I complete the glossary in the Semantic mode
- I configure the AI Agent: identity, instructions, guardrails, starter questions, skills
- I connect my real data (sources, mapping, transformations, scheduled pipeline)