Imagining the cognitive-assisted accountant


Using artificial intelligence in your accounting firm

Artificial intelligence could be part of the key to driving more focused relationships with your clients, with accountants expected to incorporate more of a data scientist element to their roles in the future.

While much of the media buzz has generated fears that new technology could do accountants out of a job, IBM’s Kieran Hagan, speaking at Accounting Business Expo 2018 in Sydney in March, says it is more likely to be about enhancing the human relationship.

Artificial intelligence and human enhancement

Artificial intelligence, or augmented intelligence, as Hagan prefers to call it, is about cognitive computing.

“It’s assisted decision-making to allow subject experts to spend more focused time with their clients while the number crunching happens in the background,” he says.

Rather than fearing the computer will dominate all aspects of service and make the services of an accountant obsolete, it’s about giving accountants better context and information to drive their decisions, recommendations and communication with clients. In addition, it can help to provide a consolidated reporting trail to use for compliance requirements. There is a range of different ways artificial intelligence could be harnessed by accountants.

Hagan provides the example of an IBM system in the US, which is utilised as a cloud-based tax assistant. It’s able to distill the 70,000 pages of annually updated US tax law into one system, cross-reference against client information ranging from images to video tapes and cross-reference key points for accountants to analyse with their clients.

Or another option might be using it in the same way Google or similar channels use data – to recognise patterns of behaviour, which might require specific services or indicate future needs. This can allow accountants to take a proactive approach with their clients, becoming data scientists to provide greater value for their clients.

The escalator to artificial intelligence

“You can’t build a brain without the information backbone,” says Hagan. Artificial intelligence needs context and statistic confidence built in for it to have value in assisting decisions.

There are three key parts to Hagan’s escalator to excellent artificial intelligence.

  1. Data: incorporating the full catalogue of information available to you.
  2. Analytics: reading the data and giving it context such that a certain combination indicates, for example, a person rather than a corporation.
  3. Machine learning: making dynamic algorithms so that the more you ask it, the more it refines itself and ensuring that simple questions can be answered.

The combination of the three should result in a system offering maximum confidence to guide decisions and recommendations for clients.

Building the right system for your business

There is a number of existing systems in the world – free and paid – that accountants can already access for their business, ranging from personality measures to tax tools.

Finding the right approach – whether existing or a new build – can be assisted using a measure known as Design Thinking, where you map out the process on a whiteboard and identify the minimum viable product to test.

Hagan says you should ask the following questions as part of the process.

  1. What do you want to do?
  2. What outcome do you want to get?
  3. What is the process and information you need to get the end outcome and what is the most effective way of reaching it?

You should also identify how to incorporate back into the moving parts of a client relationship, the training that might be required and timing for delivery.

Future thinking for artificial intelligence

Artificial intelligence and how we can use it continues to evolve at an extraordinary rate, but the tools that already exist could be revolutionary for many business models. It is important to remember that it is designed to supplement and support the human relationship – not to replace it.
To summarise, Hagan says, “Accountants should think of it as starting with the abacus, moving to the scientific calculator and now, the cognitive-assisted calculator, artificial intelligence.”

Next: What does it mean to fail fast and why you should do it often

Fail fast and fail often is a mantra oft-repeated in the world of Silicon Valley tech start-ups. And it’s one Marc Randolph firmly believes in, having applied it to the development of his best-ever idea, Netflix.

Access the latest in practice insights, market news and articles to help you grow your business.

This communication has been prepared for use by accountants only. It must not be made available to any retail client and any information in it must not be communicated to any retail client or attributed to this article. This information has been prepared by Westpac Banking Corporation ABN 33 007 457 141 AFSL & ACL 233714 (Westpac) and is current as at 27 March 2018. The information provided is general in nature and does not take into account your objectives, financial situation or needs and therefore, before acting on it, you should consider whether it is appropriate for you  It may contain material provided by third parties derived from sources believed to be accurate at its issue date. While such material is published with necessary permission, no company in the Westpac Group accepts any responsibility for the accuracy or completeness of, or endorses any such material. Except where contrary to law, we intend by this notice to exclude liability for this material. Past performance is not a reliable indicator of future performance. You should consider whether or not provision of SMSF financial advice services is appropriate for you. ©Westpac Banking Corporation 2018