The case for AI technology adoption in financial back-office roles to improve efficiency

By Tomas Gogar, AI CEO, Rossum

In this era, digital transformation isn’t anything new. Nonetheless, it can still cause a lot of confusion and resistance for some companies, many of which are often slow, unwilling or unable to implement the necessary changes to embrace technology. As a result, entire industries are barely scratching the surface when it comes to shifting to the digital world, and many, from the insurance industry to logistics and delivery are still catching up on the digital transformation.

The banking and financial sector have been notoriously slow in adapting to the online world. They paid the high price for it, giving way to a flurry of incredibly successful new disruptive players, built on cutting edge tech from the ground up. From Transferwise, Revolut or Venmo, to GoCardless, this new generation of fintech companies addressed consumers changing expectations in a way that traditional retails banks simply couldn’t.

To catch up, incumbent players have prioritised the user interfaces, giving the appearance of a digital offering, and oftentimes leaving the back end infrastructure untouched, and hence the processing power, accuracy and speed unaffected. Back-office functions, although they are essential to the smooth running of a business, have seen very little change and as a result,  too many people in these functions are still tied up typing information into spreadsheets and software forms – in fact, manual data entry is a prime example of how much resources the offline legacy wastes. Take Accounts Payable for example, invoice data entry in this sector is estimated to eat up roughly 100 human lives worth of time every single day.

With the significant increase in the number of employees working from home due to the global COVID-19 pandemic, the back-office challenges have suddenly come to light, and finally, companies that got away with minimal changes so far, are realising that they need a structural digital overhaul, and fast. We believe the solution to this is artificial intelligence backed software solutions.

Previous technology based solutions essentially did half the job, heavily depending on human fact checking. Consequently, these solutions were actually quite cumbersome and time consuming and costly to implement and maintain, and offered only incremental improvements. Now with AI, automises data processing completely removing the need for human fact checking (and human error!). Additionally, deployment is massively simplified with an average setup time of one week, compared to about 6 months for previous technologies.   AI solutions are also highly adaptable to new formats and scenarios, allowing businesses to test them in say one department and to quickly roll out a single unified solution across all functions of the business.   Data can be extracted from any invoice layout with no template or rule set-up, saving significant and effort. Rather than trying to change and standardise a highly fragmented environment (there are about as many invoice formats as there are businesses), AI can work with it, and optimise the overall process and offer a unified answer to a fragmented ecosystem.

Taking Accounts Payable as an example again, this is a sector that has relied by and large on Optical Character Recognition (OCR) software solutions in an attempt to remove some of the manual labour involved in reading processing and filing invoices. Although OCR did improve the processes to a certain degree, ultimately these types of solutions still required a long and expensive set up processes and a lot of manual labour to actually capture the data accurately with templates and manual data entry. Now, with AI software, like the one we have created, this is a solution that makes data extraction simple and easy, saving time and man power, as well as building on existing infrastructure. It has the ability to transform this industry.

In conclusion, for a sector that has been slow to adopt digital change, AI is THE technology answer that is finally fixing the invisible pain points that businesses had simply accepted as unremovable. AI applied in this way offers a viable way forward and businesses that were notoriously slow and resistant to embrace the digital transition, incentivised to make a change, may actually end up at the head of the pack. Skipping ‘older tech’ and jumping straight into AI solutions, the best scenario available by far, is indeed the smartest, fastest and most cost effective way to transition into the digital world.