The Neum AI platform tracks information across pipelines, files and users. These analytics provide high level observability as well as auditing capabilities.

Captured data includes:

  • Data processing: What was processed? Was it successful? How long did it take?
  • File management: What files are available for search? What is their status?
  • File auditing: What files are being retrieved through search?
  • User auditing: What users are searching for information? What pipelines are they accessing and what files are they accessing within?
You can opt-out of retrieval / search data capture by setting the track flag to false (default). No opt-out available for processing data captured during pipeline run.

To access the captured data, use the provided NeumClient. We have a sample Jupyter Notebook showcasing the analytics provided.

Was this page helpful?

Data retrievalLangchain