Simple ways to integrate AI into your practice

Using AI tools can save accountants and practitioners time, and add value for their clients. The key is knowing which tools to use, when to use them, and how much to rely on them. 

by | Mar 30, 2023

Woman looking at laptop

Earlier this month we looked behind the glittery curtain on the ChatGPT stage, noting that while it will be hard to avoid this next biggest thing, we can at least be aware of its shortcomings. 

Since then Microsoft has announced its equivalent: Copilot, currently in testing and to be available with all 365 paid subscriptions possibly later this year. Described by Microsoft as “an entirely new experience”, it will draw upon all a user’s Microsoft apps such as calendar, email, word and contacts to write documents, analyse spreadsheets and create plans.

It does sound too good to be true, and businesses will need to step back and decide how far they want to dive into these new pools of productivity. However, acknowledging there might be integrity, trust and sustainability credential issues that need careful management, there are several ways accounting practices can integrate AI into their systems to produce quick wins.

Make the most of what you already have

Self-described ‘Hype Girl for Accounting Apps’ Heather Smith says the first step is to ensure you’re fully optimising the features and functionalities of your existing solutions. 

“You’re likely using either the GSuite or Microsoft 365 platform, and both have built-in AI-powered functionality. The most basic way to get started is to ensure that AI is enabled,” Smith says, 

“Search the settings, and activate the AI power in everything you use. This allows you to write quickly by highlighting and correcting typos and grammatical errors as you write. It auto-suggests finishing sentences as you write them. For emails, it summarises the content and offers the email subject title; click ‘Tab’ to accept. It’s like having someone brainstorm your writing with you.

Smith suggests additional tools that go a little beyond the basics built into the platforms accountants are likely already using for client communications – including WordTune, Grammarly and ChatGPT – but recommends against over-reliance on them.

“Never use them verbatim; use them to inspire you and brainstorm with you.”

Ensure new systems use AI

Most cloud accounting solutions such as MYOB Business, QBO and Xero use AI to assist with classifying transactions, streamlining bank reconciliations, cash flow predictions and forecasts, and business performance analysis. 

“There is a plethora of new GLs [general ledgers] hitting the market; it’s worth checking to ensure they embrace AI, and offer these smart and time-saving features,” Smith says.

“Receipt-scanning solutions – think Dext, Datamolino, MMC Receipt and Squirrel Street – use AI tools to capture, extract and push data into the general ledger. 

“One of the most helpful AI-powered tools all accounting practices can easily benefit from is XBert: a data quality and workflow automation system that connects to Xero, MYOB and QBO, and analyses your files every day, multiple times a day. The AI is trained to look for anomalies, issues, or discrepancies in the books. The outcome is clean books and improved data quality.” 

Other common accounting tasks that could be handed over to AI, freeing up staff to focus on the more personalised and strategic aspects of client and business relationships, include:

  • Customer service: Chatbot providing instant answers to generic questions relating to tax rates and returns, financial reporting requirements, and accounting standards.
  • Data analysis: Running through large amounts of financial data much faster than a human could to make forecasting quicker and easier.
  • Manual data entry: Routine tasks such as data entry and bookkeeping, eliminating the chance of human error.

Be aware of the pitfalls

While some AI can seem like magic, it is only as good as the data it is drawing from and should always be checked.

PwC suggests three areas for a business to work through to check it is ready to implement AI:

  1. Define your risk appetite for new technology, which business functions you will apply it to, and how it will be monitored.
  2. Compile a thorough and clear picture of your company’s current use of AI in all forms – does it fit the guidelines you established at step one and meet other data compliance requirements?
  3. Create a governance framework to guide future development and check each aspect for compliance requirements.

Advivo Partner Chris Morris says his practice uses Dext and Hubdoc, and that these tools make a difference for the clients who are happy to use them. However, the firm hasn’t made them mandatory.

“My advice is to do your research, and that jumping in too early may not necessarily be the best use of your time,” Morris says. “Most software goes to market as a beta model and it’s essentially costing you time to teach it how it should work and iron out the kinks. We tend to sit back and see how it evolves rather than being a super early adopter.”

Embrace the benefits

There is a common perception that AI is going to replace many accounting jobs, but in a nationwide talent shortage, Smith explains, “AI and technology are effectively the first employees of many small businesses.”

Indeed, ChatGPT itself believes it cannot take the place of a human accountant. When asked, it replied:

“As an AI language model, I cannot claim to be better than a human accountant. While I can provide a range of accounting-related services, including answering questions, automating routine tasks, analysing financial data, and providing financial advice, there are certain tasks that require human expertise and judgement.”

And that’s the key to getting the best out of most AI tools – knowing what they have been trained to do well, and using them to save time and add extra value for clients by augmenting human expertise and experience.

Share This