Integrations

Every source your agents can reach. Joined in one query.

Add a catalog with a few credentials and it is queryable right away, next to everything else — databases, warehouses, lakehouses and buckets, joined in one SQL statement.

Connect your data — freeAsk about a source →
Connect everything. Query anything.
Analysts
SQL · Metabase · Superset
AI Agents
MCP
Applications
JDBC · SQLAlchemy · REST
Thunderduck Enginecost-checked, right-sized
S3 · GCS · Azure
Parquet · CSV · Excel
RDBMS
Postgres · MySQL · SQL Server
BigQuery
Data Platform
Lakehouse
Iceberg · S3 Tables

Databases

  • PostgreSQL
  • MySQL
  • Oracle
  • SQL Server
  • BigQuery
  • ClickHouse
  • SnowflakeSoon
  • TeradataSoon

Object storage

  • Amazon S3
  • Google Cloud Storage
  • Azure Blob Storage
  • Cloudflare R2

Lakehouse · Iceberg

  • AWS Glue
  • Cloudflare Basin
  • Nessie
  • Iceberg RESTSoon

On-prem · S3-compatible

  • MinIO
  • RustFS
  • SeaweedFS
  • Any S3-compatible store

Query from

  • Python · SQLAlchemy
  • Java · JDBC
  • .NET
  • REST API
  • MCP · AI agents
  • Node.jsSoon
  • GoSoon
  • AirflowSoon
How it works

From credentials to a queryable table in minutes.

1. Hand over a credential

Pick the source and enter its connection details. The connection is tested before anything is saved, and the secret goes to a secrets manager — never shown again, never in our database.

2. Tables discover themselves

Thunderduck reads the catalog's schemas, tables and columns, infers what each column means, and keeps them fresh. Agents and people browse the same map.

3. Query and join anything

Address any table as catalog.schema.table and join across sources in one statement. Filters push down to BigQuery, Postgres and MySQL; Iceberg skips files by partition and stats.

Connect your first source.

Run your first query — freeTalk to us