All core functions are available via an authenticated REST API. For seamless integration, we also offer client libraries for Java, Python and Go.
If you want to connect entire sites or machine parks to the platform, you can create and edit adapters and pipelines via the programming interfaces instead of the web-based user interface. Live and all historical data is also available via interfaces. This makes it easier to integrate external applications that are created directly by specialist departments, for example.
In short: you won't find a more open solution.
With the StreamPipes SDK, you can develop your own pipeline components - such as a driver for a proprietary fieldbus protocol, a special time series operator or a database interface.
The SDK offers a lean programming model: you describe inputs and outputs, validation logic and configuration parameters and generate a microservice with our developer tools. This registers automatically and provides the platform with new functions.
In this way, your organization's ecosystem grows with every project and reusable building blocks.
For long-running analyses, use Functions: user-defined Java or Python code that obtains all connected data streams, calculates results and makes them available again as a data stream for downstream applications.
If you want to integrate more complex ML models, use our Python integration and the integrated function zoo. Once registered, models run automatically and are integrated into the platform's metrics and monitoring system.