FAQ

Collation of frequently asked questions

Who develops Joule?

Joule is maintained by Lyndon Adams who has a extensive background in event based stream processing using distributed architectures. There is an open welcome invitation for collaborators to get involved with the project.

Why call it Joule?

A Joule is a unit of work or energy which is reflective on how the platform was envisioned and developed. A Joule process can either execute part of or complete use case depending upon complexity thus addressing multiple non-functional requirements while bringing simplicity to the solution

Why use Joule?

Low-code use case driven process, standard data connector integrations, OOTB processors, analytics first approach plus more.

Spend one hour using the platform we love to gain feedback on whats good and what needs to be improved.

Why would i use Joule verses other solutions such as Flink?

Joule has been designed to develop and pilot business use cases quickly using a low-code problem solving approach. After proving the use case there is no reason to switch to another platform but we would prefer you stay with Joule.

Do you support machine learning?

In short yes. Online ML model support is provided using the openscoring JPMML library. Supported models include Random Forest, XGBoost, K-Means, NN, regression, Bayesian etc

Do you have a SDK?

Joule ships with a SDK to enable developers to build advanced analytics, data connectors and sinks, and custom stream processors.

Can I run Joule on bare metal?

Yes, yes and hell yes. Download one of the example use case project or use the project template to get started.

Is Joule a ETL tool?

Not really. Joule wasn't designed to be a ETL tool, although it does exhibit some of the feature characteristics.

Can Joule do reverse ETL?

Yes. Joule can present domain data structures using Avro schema when using the Kafka publishing connector.