Digital transformation isn’t new, but it’s taken on a new direction with the rise of RPA. It’s more tangible than ever to see how a task can be transformed by RPA by thinking of it in simple terms: getting a robot to replicate that task for you. For example, entering values from a spreadsheet into a web form, and/or scraping customer information from one system to make it available in another.
These simple tasks within an end-to-end process, such as manual and repetitive data entry, may have come from inefficiencies created by a ‘we’ve always done it this way’ mentality and automating these tasks may not always deliver the greatest optimisation or efficiencies. True transformation, however, comes from looking at the entire process and not just replicating the actions of a human. It’s often better to revisit the entire process so it can be thought through differently (whilst being mindful to employ an agile approach to change), ensuring a valuable product is delivered as quickly as possible whilst ‘maximising the amount of work not done’.
What is possible today, however, is vertical AI: specific AI models or systems that are capable of helping us with particular tasks. If you were to give an AI model designed for customer correspondence – capable of understanding emotion, intent and subject matter – an invoice, for example, and ask it to extract the values, it’s not going to do a great job!
So, how best to make use of vertical AI? Like the tasks automated by RPA, these AI models need a place to live, they need context, and they need to be employed within a process. The distinct task they perform on their own is useless without it.
