Use cases

Start from a relationship question. The examples below show which tables to map and what to inspect in the result.

On this page

Transaction investigation

Question. Which accounts receive funds two transfers away from a starting account?

Map accounts to Account nodes and transfers to Transfer edges with amount and event time. Use tenant-aware keys. Rank destinations by path count or amount, then inspect individual transfer paths and timestamps.

Cypher
MATCH (a:Account)-[:Transfer]->(b:Account)-[t:Transfer]->(c:Account)
WHERE a.tenant_id = 1 AND a.account_id = 10
RETURN c.account_id AS account, count(*) AS paths, sum(t.amount) AS amount

Run this example on the included Parquet dataset.

Service impact analysis

Question. Which services depend on the component being changed?

Map an inventory table to Service nodes and calls or declared dependencies to DEPENDS_ON edges, directed from consumer to dependency. Keep ownership and environment as properties. For this illustrative schema, the query follows incoming dependencies up to three hops from payments.

Cypher
MATCH p = (s:Service)-[:DEPENDS_ON*1..3]->(target:Service)
WHERE target.name = 'payments'
RETURN p LIMIT 100

See how to map node and relationship keys before running this on your own inventory.

Data lineage

Question. Which upstream datasets contributed to a report table?

Map catalog entries to Dataset nodes and lineage records to DERIVED_FROM edges, directed from output to input. Add job or pipeline identifiers as relationship properties. For this illustrative schema, explore paths up to four hops from the report.

Cypher
MATCH p = (report:Dataset)-[:DERIVED_FROM*1..4]->(:Dataset)
WHERE report.name = 'daily_revenue'
RETURN p LIMIT 100

Use the graph view to inspect paths and share findings with dataset owners.