Building a Centre of Excellence #8 - Stage 3 : Process Definition Document
With the process deep dive completed and having a recorded video of the process, it is now time to create the process definition document (PDD). This document is critical and acts as an insurance policy for establishing a contract of delivery with our internal customers.
The PDD is a step-by-step guide to every screen, keystroke, mouse click and data handling and manipulation. For the first pass through we take high definition screenshots from the Teams recording and add commentary to each stage.

Creating the PDD can be a very time-consuming process, but it is so important to get it right. The PDD will become the programming bible for the bots, and their accuracy and success depends on this being correct.
For the second pass, we spend time reflecting on each stage and thinking what could go wrong, for example, a duplicate patient, patient record not found, input data in wrong format etc.
Any potential error you can double guess can be built into an exception at the very outset meaning you can teach the bots to handle these eventualities. Identifying potential issues now will result in a lower exception rate at the point of go-live.

The third stage, and here is the golden egg, is the critical difference and why ESNEFT has been successful. We now take the PDD, give it to someone who doesn't work in the target department and have them go through the PDD step by step. If we get the correct output at the end, then we know we have mapped the process 100% accurate.
This stage is also beneficial for identifying any subconscious decision making. Very often, we find that the PDD fails as the originating member of staff makes critical decisions without even thinking about it. Very similar to driving the car and not thinking about changing gear or route you have just taken. We now have the opportunity to understand this subconscious decision making and map it correctly.

Finally, we create a high-level data schematic to ensure we have captured all of our data needs (see example above)