Cross-source analytics
Join your managed PostgreSQL with your Iceberg tables in a single query
Multiple sources, one SQL query
Federated lakehouse
All your sources, one SQL query.
Break down organizational silos
Transform your scattered data sources into unified analytics power. Built on Trino and Apache Iceberg, Data Hub federates your managed PostgreSQL and your Iceberg tables on S3 object storage into a single queryable base, and is fed through the Open Data, Google Drive and Email connectors.
Specifications
- Federation
- Managed PostgreSQL, Iceberg on S3
- SQL Engine
- Distributed Trino
- Ingestion
- Open Data, Google Drive, Email
- Cross-source
- Multi-source joins
- Storage
- S3 buckets created from the console
- Medallion
- Bronze, silver, gold Iceberg schemas
- Dataset Branching
- Roadmap
Use cases
Complex KPI calculation
Advanced analytics on distributed datasets
No ETL pipeline to maintain
Unified lakehouse
Open data into Iceberg, Google Drive files and email attachments into S3: connectors feed your queryable lakehouse
No ingestion code to write
In action
The impossible join
An analyst needs to cross customer data stored in PostgreSQL with product events stored in Iceberg tables
- Declare the managed PostgreSQL catalog and the Iceberg catalog in the SQL engine
- Write a single SQL query joining both sources
- SELECT * FROM postgres.customers JOIN iceberg.events...
- Read the results without copying or moving the data
A multi-source join with no prior extraction and no intermediate copy
Consolidated reporting
A finance team consolidates its monthly reporting from its PostgreSQL database and its Iceberg tables
- Federate the managed PostgreSQL database and the Iceberg catalog in the Trino SQL engine
- Write a query covering both sources in the SQL editor
- The Trino engine distributes the computation across the cluster
- Save the query and replay it at every close
Consolidated reporting, queryable in SQL, with no pipeline to maintain
A lakehouse fed from the console
A data team feeds its lakehouse with open data datasets and collects its business files from Google Drive and an email inbox
- Create the lakehouse S3 buckets and an Iceberg catalog from the console
- Import an open data source into an Iceberg table of that catalog, one off or daily
- Connect the Google Drive and IMAP email connectors to drop files and attachments into the buckets
- Query the Iceberg tables in standard SQL from the console editor
A lakehouse fed from the console, with no ingestion code to write
Powered by open standards
SQL engine
Trino
Distributed and stateless SQL engine. Query your connected sources with standard SQL, without moving your data.
Table format
Apache Iceberg
Open source table format for data lakes: ACID transactions, schema evolution, compaction and snapshot expiry driven from the console.
Connectors
Multi-source federation
- PostgreSQL
- Managed database, read only
- Apache Iceberg
- Lakehouse tables
- S3
- Object storage, exposed through Iceberg
- MySQL
- Relational databases (coming soon)
- MongoDB
- NoSQL (coming soon)
- Elasticsearch
- Search (coming soon)
Ingestion connectors
- Open Data
- 23 public sources, one off or daily ingestion into your Iceberg tables
- Google Drive
- File synchronization into your S3 buckets, with optional text extraction
- Attachments from an IMAP mailbox dropped into your S3 buckets
Open data, files and emails feed your Iceberg tables and S3 buckets with no pipeline to write. More federation sources are coming soon.
Key benefits
- Query your connected sources with standard SQL
- No ETL required: analyze in place
- Distributed computation for massive scale
- Apache Iceberg for ACID transactions
Ready to unify your data sources?
Discover how Data Hub can eliminate your data silos.