I'm looking for the job management/running alternative, not the algorithms etc.

Our current system uses Elastic multi-metric jobs to detect anomalies over thousands of metrics with zero effort - we are able to add additional metrics to the feed, which are automatically enrolled.

To be clear, this is not multi-variate anomaly detection - this is applying the same detection algorithm on many separate and distinct series in order to process a lot of series at scale to detect anomalies (outliers, missing data, etc) in each single, unrelated series, but doing this in bulk rather than configuring one job per series.

With elastic you feed it a dataset (via an elastic query) and provide a result property to split on, creating the multiple metrics - these effectively become their own model, except they're all processed in a single managed job.

The Elastic machine learning is fantastic for this purpose, and I struggle to find an alternative, which I'm looking for because of $$.

Thanks.