Why there is a step-change in automation of core business processes?

I think basically because it has become so accessible and so cheap.

It’s quite a mind change.

While there is understandably much attention to the magic of AI and machine learning, I would not be surprised if there is much more real world impact of its less glamorous and older sister, Robotic Process Automation (RPA).

RPA in the world of government and business processes is about software that does repetitive information processing work on a computer. Repetitive work that humans were doing before, possibly on paper, but mostly by pulling data from one system, cutting and pasting into another, populating data fields and completing forms.

With RPA, we can ‘train’ a software to do the same. Open and close different programs, copy fields, paste, and edit text. Often, they can be programmed through a visual interface on the screen. They are mostly executed an analyst on a laptop.

The business case is usually that there is much such repetitive work across a government or business. Often this is done to compensate for small problems with the core business systems. Perhaps there are some missing information fields in the core system, or a need to transfer information between i.e an HR system and a payments processing system.

Tedious work really. RPA relieves humans of repetitive tasks, and also helps avoid the kind of mistakes, i.e typos, that often occur when humans do the same.

It has become attractive for several reasons. Most importantly, as it allows for a simple way of automating tasks without rebuilding the big corporate and legacy systems. It’s cheap and with very quick return on investments.

IT-developers tend to prioritize the bigger issues, and not the smaller business problems. Front-end users might find that instead of negotiating a backlog of issues with the IT-department, the front-end users can themselves program automation between different softwares through very low barrier interfaces. The RPA solutions are practically programmable by any smart business analyst.

They require no infrastructure, expensive servers and physical things. Cost are subscription-based, and very low compared to traditional IT projects. The most substantial cost is at the start-up, where there is a learning curve, for analysts and managers, until the organization figures out how to best take this to scale.

The business case makes even more sense when applied to a set of related workflows or processes. The marginal costs of adding new processes are very low, and it falls as programming skills in the organization further improves.

Hence the revolution.