Volume and timings
- Danielle Rubas (Deactivated)
Volume
Volume is arguably the easiest chart. Here you can see a time series of the number of documents that went through the system and the number of documents that has been finalised.
You can choose to see the incremental view (absolute volume per day) or the cumulative view (volume processed to date).
Efficiency gains
The raison d'être of our solution is that we save time. On the one hand, our solution can be used for your end customers. To help them save time when they onboard, when they exchange information, ... On the other hand, our solution can be used by your teams to save time in routing of incoming communications, in data entry tasks, ...
As such, statistics around timing are very important. Similarly, we track the number of keystrokes saved by our solution.
Average time
We allow you to see statistics between any two of the five above timestamps. We track five timestamps
Document queued
Document processing started
Document processing done
Document review started
Document review done
Volume | In this graph, you can see the volume of documents received and reviewed on a daily basis. |
Average time | The average time between two timestamps. The five timestamps (1) received: when the document was uploaded, (2) started: when we started to process the document (for asynchronous calls it starts in a queue), (3) completed: when we have processed the document, (4) review_started: when the user started to review the document, (5) reviewed: when the user finalised the review of the document |
Average time spent by reviewer per document | The average time spent by a reviewer per document. Note that this is the actual time spent and not the clock time (if the user worked on it during two sessions, the time of these sessions will be added up, not the time from the start of the first session to the end of the last session.) |
Average time spent by reviewer per field | The average time spent by reviewer per document divided by the number of fields. |
Label frequency | The number of times a label has been seen. Typically, you will look at the difference between the label frequency in gold and predicted. |