Auto-response retrieves the processed results for the n-th request during the n+1-st request.This allows for almost real time data retrieval. For example, say you are processing constantly updating data such as a Twitter stream. On Second 1, you submit documents 1 to 10. Since this is your first submission of the day, no documents are returned to you. On Second 2, you submit documents 11 to 20. This time you will receive documents 1-10 back in the submission response. On Second 3, you submit documents 21-30, and you get back documents 11-20, and so on.

If you are processing a bounded amount of content, such as a set of 1,000 survey responses or the like, auto-response is not a good choice, as you need to remember to make another call at the end to get the last submitted batch. Auto-response is really designed for a constantly updating stream of data with no beginning and no end.

Auto-response is disabled by default and is enabled per-configuration. If there is no processed data, Semantria will respond with a 202 HTTP status.

Semantria will respond with an array of up to n processed documents following the status message. The default number of documents is 2 for almost real-time speed. For high volume, we suggest 100 or more documents. Please contact us to make a change.



  • Fastest retrieval option: almost real time
  • Automatic: does not require dedicated data retrieval job to get processed data back


  • Batch size and sequence not preserved in retrieval
  • Does not allow for retrieving large amounts of processed data at once

Use Cases

  • Users interested in 24/7 data processing and continuous usage
  • Users who want to retrieve data as fast as possible
  • Those who want to monitor social media trends
  • Business intelligence and trending services