Every source, woven into one living graph.
Connectors sync your sources in place; auto-discovery maps them; entity resolution weaves them into a governed knowledge graph — virtualized data structures the whole organisation can query, govern and build on.
IT developers
Create and consume APIs to bring new technologies and data types into the fabric.
Citizen developers
Build insights with the data they’re granted, inside business applications.
Consumers
Employees, customers and trusted partners access data under universal governance.
The fabric’s output: an embedded semantic layer.
Every source speaks its own schema, and “revenue” or “active customer” means something different in each. The Knowledge Fabric extracts those schemas automatically and resolves them into one overall ontology — an embedded semantic layer that fixes what your data means, once, inside your environment, and serves that meaning to every dashboard, app and agent.
One agreed meaning for every metric, entity and relationship.
Meaning is context. When definitions are pre-resolved in the fabric, agents reason less, spend fewer tokens, run faster — and make fewer mistakes.
Schemas, read for you
Connectors and auto-discovery read every source’s schema — fields, types, keys, relationships — and keep it in sync. No hand-built models to maintain per system.
Resolved into meaning
Entity resolution unifies the extracted schemas into a single ontology: one customer is one customer, “revenue” has one agreed definition, and the links between them are explicit.
Served everywhere — and yours
The semantic layer is embedded in the fabric, inside your environment — not rebuilt in every warehouse, lakehouse and BI tool, and not locked in someone else’s cloud. Define meaning once; serve it over SQL, GraphQL and MCP.
A defensible core, in concentric layers.
A defensible core, in concentric layers.
FabriCloud Graph is the Knowledge Fabric at the centre of the platform — components, output and ecosystem wrapped around a single core, with the spectrum of agent autonomy running beneath it.
Fabric
Fabric Components
Fabric Output
Fabric Ecosystem
Agent types & AI autonomy
FabriCloud runs agents at five autonomy levels — from fully deterministic (0) to fully autonomous reasoning (4). Each level trades human determinism for AI autonomy.
Deterministic Agents
Follow predefined, fixed steps. No reasoning autonomy — fully predictable and repeatable.
Human-in-the-Middle
ReAct reasoning with a human in the loop — AI proposes, a person reviews and approves at each decision point.
BPM-Governed Agents
Autonomy bounded by a defined BPM process — the agent acts freely within the steps and gates of a modelled business process.
Self-Organising Regulated Agents
DAG agents that self-organise from user-defined rules & dependencies — plan their own path, escalate on exceptions.
Pure Agentic
Fully autonomous ReAct — sets its own goals and steps end to end, with post-hoc audit.
The engine that turns data into context.
Most platforms move and store data. The Knowledge Fabric understands it — resolving entities, mapping relationships and carrying meaning, so search, enrichment and AI work on context, not disconnected rows.
Entity resolution
Match and merge records that point to the same real-world entity across every source.
Knowledge graph
Turn entities and relationships into a queryable graph — connections become first-class.
Metadata catalog
A living map of what you hold, where it lives and what it means.
Lineage & governance
Every transformation tracked, every access governed — trust travels with the data.
Resolve who an entity is — from how it’s connected.
Attribute matching finds the obvious duplicates. Graph entity resolution finds the entity you’d miss — and refuses the merge you’d regret. Records with different attributes become one entity when their relationships overlap; one shared name splits into two when the graphs don’t. It resolves and disambiguates by connections, not strings — and every decision is explainable by the path through the graph.
And resolution isn’t a step before the graph — it is the graph. Every resolved entity is instantly queryable, in context, by everything the platform runs.
Bring the cloud to your data.
See it deployed in your own environment — analytics, apps and agents on governed data — a boundary raw data never crosses.
Start building