For many of the use cases that our solution is used for efficiency is very important.
Efficiency is basically a function of how good our ML is at predicting the right answer (no human intervention required) and how fast people can complete the predictions of the ML algorithms.
To allow people to quickly review documents, we have introduced two key concepts:
- Exception-based review: This allow you to only look at places where there is low confidence. You can alway jump to the next thing to review simply by pressing ENTER
- Pick next: We will pre-load the data for the next document to review while you are reviewing a current document so you are not delayed by the network