Skip to content

Connect a Database

Choose the source that contains the answer you need. If you already have approved connection details, open its guide below and enter them in Add connection. Otherwise ask the owner for a reporting identity, the relevant tables or views, authentication details and a network path from Datarelix.

A sign-in account for Datarelix is separate from permission to read a database. Each conversation and dashboard uses one connection. Snowflake uses password-only authentication here; check its current account policy before setup.

Supported dialects

DialectQuery languageDefault scopeAuth modes
PostgresSQLpublic schemaUsername & password · Connection string
MySQLSQLconnected databaseUsername & password · Connection string
Azure SQL / SynapseT-SQLdbo schemaUsername & password · Connection string · Entra OBO · Service principal — secret · Service principal — certificate
DatabricksSpark SQLdefault schemaPersonal Access Token · OAuth M2M · Federated JWT
Kusto (Azure Data Explorer)KQLconfigured databaseEntra OBO · App + Certificate
BigQueryGoogleSQLpick a datasetService account JSON · OAuth
AthenaTrino/Presto SQLpick a Glue databaseIAM access keys · Assume Role
ElasticsearchES|QL (8.11+)pick allowed indicesAPI key · Basic auth
SnowflakeSQLconfigured schemaUsername & password
Files & Object Storage (DuckDB)DuckDB SQLconfigured datasetsAWS S3 · Azure Blob/ADLS · Google Cloud Storage · Direct HTTPS

Common fields

Every connection has these:

  • Name — your label for this connection.
  • Dialect — pick one of the dialects above.
  • Allowed scope — Datarelix only reads structure and runs queries within this scope. The label changes per dialect (schema / database / dataset / index pattern), but the meaning is the same: the configured boundary for discovery and reporting queries; also restrict the source identity’s effective permissions.

Dialect-specific fields appear after you pick a dialect. See each per-dialect guide for details.

After you connect

Datarelix runs discovery: it reads the tables, columns, types and declared keys in the allowed scope. You can then run AI schema analysis, which adds descriptions and suggests relationships, plus primary keys for tables that declare none. AI schema analysis counts against your plan’s monthly allowance, and you can skip it.

Before analysis starts, you can add optional context: what the database contains, which tables matter most, which columns act as business keys. These notes are sent with the schema structure; no row data is included in discovery or AI schema analysis.

Edits you make on the schema page — descriptions, keys and relationships — take precedence over what discovery and AI schema analysis found. They only change how Datarelix interprets questions; they never change your database.

Re-run discovery any time you add columns or tables. Review the descriptions and relationships before relying on them in a question.

Universal troubleshooting

SymptomLikely causeFix
Connection refusedNetwork reachabilityAllow Datarelix’s egress IPs in your database firewall.
Authentication failedWrong credentialsTest with the dialect’s native CLI or console first.
Empty schemaWrong scopeCheck Allowed scope — see the per-dialect guide for what blank means.
AI schema analysis takes too longA very large schemaNarrow the allowed scope, or skip AI schema analysis and use discovery only.

For the error messages shown in the app, see Troubleshooting.