The DSS API Node makes it easy to expose predictive models through a REST API.
By using DSS only, you can compute predictions for all records of an unlabeled datasets. Using the REST API of the DSS API node, you can request predictions for new previously-unseen records in real time.
The DSS API node provides high availability and scalability for scoring of records.
It can expose as API both:
- “Regular” prediction models, trained using the DSS machine learning component
- “Custom” prediction models, written in Python.
Thanks to its advanced features, the DSS API node is at the heart of the feedback and improvement loop of your predictive models:
- Powerful logging and auditing capabilities
- A/B testing and multi-version evaluation of models
- User-aware version dispatch
For more information about the user-API, see API node user API
Creation and preparation of models used by the API node is always done using DSS. The API node itself is a server application-only, it does not have an UI. You manage the API node through a REST API or a command-line tool. See Using the apinode-admin tool and API node administration API.