Find the relationships your team forgot to document
Your databases have more relationships than anyone documented. Squish scans your schemas and actual data patterns to surface connections that exist only as tribal knowledge.
What makes Squish different
Not another ERD tool
Traditional tools show you what you already know. Squish discovers what nobody documented.
Cross-Database Discovery
Find relationships that span PostgreSQL, Snowflake, BigQuery, MongoDB, and more. When user_id in your orders service connects to user_id in billing, we find it -- across SQL and NoSQL.
Statistical Analysis
We analyze actual data patterns, not just column names. UUID matching, cardinality analysis, and value overlap detection surface hidden connections.
Explainable Results
Every relationship comes with evidence: why we think this connection exists, what data supported it, and how confident we are.
Four levels of discovery
How we find relationships others miss
Constraint Extraction
Extract declared foreign keys, primary keys, unique constraints, and indexes from your database metadata.
Column Inference
Match columns by naming patterns. When user_id appears in multiple tables, we infer the relationship across snake_case variants.
Statistical Analysis
Sample actual data values, compute overlap statistics, detect UUID patterns, and analyze cardinality to find hidden connections.
NoSQL Pattern Analysis
Detect MongoDB embedding patterns, DynamoDB GSI relationships, Cassandra partition key structures, and cross-reference them with your SQL schemas.
Squish vs. the status quo
Stop sending Slack messages asking which tables connect
Connect your databases and see relationships in minutes. Read-only access, configurable confidence thresholds, explainable results.