Lakehousecat connects to relational SQL databases via a connection URI. Once connected, the semantic layer is built from the database schema, enabling the AI to query your tables using natural language.
Relational & Cloud Databases
| Database | Connection URI Prefix | SSH Tunnel |
|---|
| PostgreSQL | postgresql:// | ✅ |
| MySQL | mysql:// | ✅ |
| MariaDB | mariadb:// | ✅ |
| Microsoft SQL Server — driver via job definition | mssql:// | ✅ |
| Azure Synapse Analytics — driver via job definition | azuresynapse:// | ✅ |
| Oracle Database | oracle+oracledb:// | ✅ |
IBM DB2 — amd64 only, driver via job definition | db2+ibm_db:// | ✅ |
| Google BigQuery | bigquery:// | ✅ |
| Amazon Redshift | redshift+psycopg2:// | ✅ |
| Amazon Athena | awsathena+rest:// | ✅ |
| Snowflake | snowflake:// | ✅ |
| Databricks | databricks:// | ✅ |
| Databend | databend:// | ✅ |
| Vertica | vertica:// | ✅ |
| Exasol | exasol:// | ✅ |
| MonetDB | monetdb:// | ✅ |
| SingleStore | singlestoredb:// | ✅ |
| Greenplum | greenplum:// | ✅ |
| PolarDB | polardb:// | ✅ |
Distributed & NewSQL
| Database | Connection URI Prefix | SSH Tunnel |
|---|
| CockroachDB | cockroachdb:// | ✅ |
| YugabyteDB | yugabytedb:// | ✅ |
| Vitess | vitess:// | ✅ |
| TiDB | tidb:// | ✅ |
| OceanBase | oceanbase:// | ✅ |
Time-Series Databases
| Database | Connection URI Prefix | SSH Tunnel |
|---|
| TimescaleDB | timescaledb:// | ✅ |
| QuestDB | questdb:// | ✅ |
Streaming & Real-Time
| Database | Connection URI Prefix | SSH Tunnel |
|---|
| Materialize | materialize:// | ✅ |
| CrateDB | cratedb:// | ✅ |
| RisingWave | risingwave:// | ✅ |
| ClickHouse | clickhouse:// | — |
In-Process & Embedded
| Database | Connection Method | SSH Tunnel |
|---|
| DuckDB | File upload | — |
| SQLite | File upload | — |
NoSQL Databases
| Database | Connection Method | SSH Tunnel |
|---|
| MongoDB | mongodb:// | — |
| Redis | Host / Port | — |
Query Engines & OLAP
| Database | Connection URI Prefix | SSH Tunnel |
|---|
| Apache Drill | drill+sadrill:// | ✅ |
| Apache Druid | druid+pydruid:// | ✅ |
| Apache Hive | hive+pyhive:// | ✅ |
| Trino | trino:// | ✅ |
| Presto | presto:// | ✅ |
| Apache Doris | doris:// | ✅ |
| StarRocks | starrocks:// | ✅ |
| Apache Spark SQL | sparksql:// | ✅ |
| AlloyDB Omni | alloydb:// | ✅ |
| HydraDB | hydradb:// | ✅ |
Connection Requirements
- Microsoft SQL Server, Azure Synapse and IBM DB2 need a vendor driver that is provisioned once through a job definition — see Vendor Drivers
- A read-only database user is recommended
- The user must have SELECT access to the schemas and tables you intend to expose
- Network connectivity from the Lakehousecat backend to the database — direct or via SSH tunnel
SSH Tunneling
Most database types support connecting through an SSH bastion host. Enable the Use SSH toggle in the creation form to reveal the SSH settings fields.
See SSH Tunneling for full configuration details.
Next Steps
See the individual type pages for connection URI examples, or refer to the Data guide for the full creation workflow.