The Endeca ITL uses an instance configuration to process, tag, and locate data.
Creating an instance configuration is an iterative process. Endeca recommends that you first create a very simple instance configuration to test your data. After the simple configuration is working as you expect, you can make additional modifications, view your results, and make changes as necessary. Also, it is often useful to work on a subset of your data, for quicker turnaround of data processing, while you are developing your instance configuration.
At a high level, Endeca ITL development looks like this: