← All capabilities
Machine Learning
Classic ML is still in the platform, now as tools agents and ETL jobs can call, not as the only product story.
Train regression, classification, clustering, and anomaly models against a DSN source. No cluster to provision; the API owns the job.
Pipeline model steps and LangGraph tools use the same endpoints, so a nightly ETL can train and a supervisor can score in the same catalog.
What this does for the business
- Stand up a baseline model before a data-science sprint is staffed.
- Score new batches as they land from an ETL run.
- Compare clustering methods without standing up notebooks in production.
API
POST /api/models/predict POST /api/models/kmeans Full reference: api.dsnresearch.com/docs