Graph Database connector

The Neo4j data catalog your whole team can trust.

The Dawiso Neo4j data catalog turns the property graph into a searchable inventory: every node label, relationship type and property, with links to relational and BI catalogs.

Read the docs

Live connector Stable connector

Neo4j

Dawiso

Metadata-only · your data never leaves the source

Type: Native graph database
Auth: Username & password (read-only role)
Sync: Scheduled, incremental
Direction: Read-only · metadata

First things first

What is a data connector?

Metadata-only Read-only access Incremental sync Cross-system lineage

A data connector is the bridge between a tool in your stack and the catalog that gives you a unified view of it. Once a connector is configured, it reaches into the source system on a schedule, reads out the metadata - schemas, tables, dashboards, jobs, ownership, lineage - and represents it inside the catalog. Your actual rows and values stay where they are.

Connectors are the reason a data catalog can answer questions like "which Power BI dashboard depends on this Snowflake table?" or "who owns the orders topic in Kafka?" - automatically, without anyone keeping a spreadsheet up to date.

Three properties separate a good connector from a brittle one: it should be read-only and safe, it should be incremental so a full re-scan isn't required for every refresh, and it should resolve lineage across system boundaries, not just inside one tool.

About the platform

What is Neo4j?

Neo4j is the native graph database founded in Sweden in 2007 and the most widely deployed graph database according to industry rankings. It stores data as nodes, relationships and properties, runs on the Aura managed service or self-hosted, and underpins everything from the Panama Papers analysis to NASA project graphs and GraphRAG knowledge graphs.

Graphs are powerful precisely because the schema is implicit: nodes carry labels, relationships carry types, properties appear and disappear. That's a feature for modelling, a problem for governance: there is no information_schema for property graphs. That's where the Dawiso Neo4j data catalog joins the picture: read-only, metadata-only, and cross-platform.

Architecture

How Dawiso connects to Neo4j

A small read-only role on the Neo4j side. The Dawiso scanner pulls metadata on a schedule. Everything ends up in your catalog, business-readable.

Source

Neo4j database

Databases (Enterprise multi-DB)
Node labels & properties
Relationship types
Indexes & constraints

REST · JDBC

Dawiso scanner

Read-only metadata

Schema & object discovery
Dependency resolution
SQL flow parsing (optional)
Sampling on opt-in

Internal

Catalog

Dawiso platform

Searchable metadata
Lineage & ownership
Business glossary
Policy & classifications

Connection details

Protocol: Bolt protocol (Cypher) · SHOW INDEXES / CONSTRAINTS / PROCEDURES
Authentication: Dedicated role with MATCH, TRAVERSE and SHOW privileges
Lineage: Neo4j does not expose native lineage; Dawiso documents the property graph schema (node labels, relationship types, property keys) and links graph entities to upstream relational sources at the data-product layer

Setup

Connect Neo4j in 4 steps

01

Switch to the system database

Connect as a Neo4j administrator and run :use system. On Community Edition role and user creation is not supported; use the default neo4j user. Aura requires role management in the Aura console.
02

Create a catalog-reader role

CREATE ROLE dawiso_catalog_reader; then grant per target database: GRANT ACCESS, GRANT MATCH {*} ON GRAPH, GRANT TRAVERSE, GRANT SHOW INDEXES / CONSTRAINTS / PROCEDURES / FUNCTIONS and GRANT EXECUTE PROCEDURE db.info.
03

Create the user and assign the role

CREATE USER dawiso_reader SET PASSWORD '...' CHANGE NOT REQUIRED; GRANT ROLE dawiso_catalog_reader TO dawiso_reader. Read-only end to end.
04

Connect via Bolt and ingest

Provide the Neo4j connection URL (neo4j+s:// for Aura, neo4j:// for self-hosted), database name, user and password. Scheduled incremental sync from there.

Capabilities

What you get with the Neo4j connector

Property graph schema

Every node label, relationship type and property key is catalogued. Inferred types per property, plus the indexes and constraints that govern them.
Node & relationship documentation

Document what a (:Customer)-[:OWNS]->(:Account) relationship means in business terms, alongside the equivalent customer table from the relational source.
Cross-platform graph lineage

Link Neo4j node labels to upstream relational tables and downstream BI dashboards so the graph stops being an island in the catalog.
PII classification

Classify a node-property pair once. Dawiso flags every Neo4j property carrying email, IBAN or government IDs across every database and instance.
Business glossary alignment

Tie node labels and relationship types to glossary terms so a graph engineer and a data analyst speak the same language about (:Customer).
Read-only by design

Dawiso reads through a role with MATCH, TRAVERSE and SHOW privileges. No write privileges anywhere, no modifications to the graph.

Business value

Why teams turn on the Neo4j connector

−65%
Fewer 'what does this label mean?' pings

Engineers find the documented node label and its relationships in Dawiso instead of asking the graph team what (:Customer)-[:OWNS]-> actually points to.
End-to-end
Lineage past the graph boundary

Connect Neo4j node labels to the relational tables that feed them and the BI reports that consume them. Graphs stop being a catalog blind spot.
Audit-ready
GDPR & DORA evidence

Sensitive node properties are classified once, the policy follows them across the property graph and into downstream catalogs, with a full audit trail.

Ready to catalog your Neo4j?

Set up the connector in an afternoon. See your first lineage graph the same day.

Read the docs

Frequently asked questions

Still curious? Talk to our team ->

What is a metadata graph?

A metadata graph models the relationships between data assets - tables, columns, jobs and owners - as a graph. Dawiso reads Neo4j metadata read-only and catalogs labels, relationship types and properties, linking your graph data to the rest of your stack.

Does Neo4j have a data catalog?

No native catalog. Dawiso reads Neo4j schema read-only - node labels, relationships and properties - and turns it into a searchable catalog with ownership, documentation and cross-platform lineage.

What permissions does Dawiso need in Neo4j?

A dedicated role (e.g. dawiso_catalog_reader) with GRANT ACCESS, GRANT MATCH {*} ON GRAPH, GRANT TRAVERSE, plus GRANT SHOW INDEXES / CONSTRAINTS / PROCEDURES / FUNCTIONS and GRANT EXECUTE PROCEDURE db.info. Assigned to a user with CHANGE NOT REQUIRED. Read-only.

Does Dawiso copy our Neo4j data?

No. Dawiso reads the property graph schema, indexes and constraints. Node and relationship metadata is catalogued; property values are not copied. Sampling for type inference is opt-in per data source.

Aura, self-hosted, Community Edition - which are supported?

Enterprise and Aura both, with role-based access control. Community Edition does not support CREATE ROLE / CREATE USER, so connect with the default neo4j user. Aura connections use neo4j+s://; self-hosted uses neo4j://.

How does graph lineage work in Dawiso?

Neo4j does not expose native lineage. Dawiso catalogues the property graph schema (labels, relationship types, properties) and links graph entities to their upstream relational tables and downstream BI consumers through the data-product layer.