Intel / article

Real-Time Intelligence with IBM Time Series Models on Confluent

Source: huggingface.co
Date: Published Sep 2, 2026

Context

IBM's time-series models bring forecasting into Confluent Cloud's live data pipelines through an early-access service. A time series is a sequence of measurements over time. Teams can pass those measurements to forecasting or anomaly-detection functions written in Flink SQL, the query language of Apache Flink, a stream-processing engine, while the platform manages model serving. IBM Research's announcement.

The integration also handles the history and delivery around each prediction. Flink maintains recent measurements for each series, because a forecast only means something against recent history. results are written to Kafka topics, named streams of events that alerting systems, dashboards and other applications can consume. That connects an incoming measurement to a prediction and its downstream use within one workflow. Integration details.

Terms in this piece · Glossary

inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
streaming — Sending a model's response token by token as it is generated, so the reader sees text immediately instead of waiting for the whole answer.

Read the source huggingface.co