Kettle is the open-source ETL engine better known today as Pentaho Data Integration (PDI). The name is a backronym - "Kettle E.T.T.L. Environment" - and it stuck with a generation of data engineers who built pipelines in its Spoon visual designer. Now maintained under Hitachi Vantara, PDI remains a mature, capable ETL tool with a free community edition and a commercial enterprise tier, and it is still in wide use across on-prem and hybrid data shops.
Transformations and jobs
PDI organizes work into two levels. Transformations move and reshape data row by row - reading from sources, applying steps like joins, lookups, filters, and calculations, and writing to destinations. Jobs orchestrate the bigger picture: running transformations in sequence, handling conditions, moving files, sending notifications, and managing dependencies. You build both visually in Spoon, and run them from the GUI, the command line (Pan and Kitchen), or a scheduler.
Why it endures
PDI's longevity comes from being genuinely free to start, broadly connected (databases, files, big-data sources, APIs), and not tied to any one cloud. For teams with on-prem databases, mixed environments, or a preference for self-hosted tooling, it offers full-featured ETL without a per-row bill. A large body of existing pipelines and community knowledge keeps it relevant even as newer ELT tools take the spotlight.
Where it fits
- On-prem and hybrid environments where data shouldn't or can't move to a managed cloud service.
- Teams wanting full-featured, visual ETL with no licensing cost to begin.
- Organizations already invested in the Pentaho/Hitachi stack for analytics and reporting.
When to look elsewhere
For cloud-warehouse-centric work, modern ELT tools like Fivetran or Matillion fit the "load then transform in the warehouse" pattern more naturally. Teams that want pipelines as version-controlled code rather than visual jobs, or that need elastic cloud scaling, may find PDI's desktop-rooted model less aligned with current practice.
How it compares
Kettle/PDI sits alongside Talend as one of the two classic open-source visual ETL tools - both mature, both visual, both Java-rooted. Where Talend generated standalone code and broadened into data quality, PDI runs transformations on its own engine and pairs tightly with the Pentaho BI suite. Against cloud-native ELT, the trade is self-hosted control and zero per-row cost versus managed convenience.
Frequently asked questions
Is Kettle the same as Pentaho Data Integration? Yes. Kettle is the original project name; Pentaho Data Integration (PDI) is the product it became, now under Hitachi Vantara.
Is Pentaho Data Integration free? There is a free, open-source community edition alongside a paid enterprise edition with added support and features.
What is Spoon in PDI? Spoon is the graphical designer where you build transformations and jobs; Pan and Kitchen run them from the command line.