One of the other hats I wear is that of a trainer. With 20 years of experience and an accounting audience in mind, I carefully craft the titles, synopses and learning outcomes, and fully develop training sessions.

Recently, a client processed my training sessions through generative AI and the cheesy, used-car-salesman outcomes were completely different from what I’d initially submitted. For example, the copy now mentioned how the session would help you do your taxes, even though the original session had nothing to do with taxes! To compound this, the client then queried why the agenda didn’t align with the title, which the AI had changed!

While I fully embrace the magic of AI, I was upset. The client’s processes missed engaging a human to quality-check the outcome and they disrespected the years of experience I bring to the table.

AI vs automation: What’s the difference?

The term ‘artificial intelligence’ was coined in the 1950s by Alan Turing in his work Computing Machinery and Intelligence. It’s undergone multiple evolutions since then.

AI is not automation: the integration of workflows across different solutions. In simplistic terms, AI can be understood to be algorithms and datasets. Accountants understand algorithms, and accountants have access to lots of datasets. We have our own data, our individual and aggregated client data. Given a push, we understand macro- and micro-economic data. There’s lots of data we can overlay with AI magic.

“Sometimes, I will put in something personal, like I’ll tell them a story about my dog, 'Chester, my border collie, is hoping I can sort this out quickly so I can take him for a walk', to help figure out if I am speaking to an actual human. Who can resist talking about a cute dog?”

Navigating AI-driven customer service

I personally think the most important deciding factor in choosing technology is the support the company offers. One of AI’s most prolific business uses has been within customer support.

Businesses quickly tried to enhance efficiencies and save money by replacing humans with AI responses. Unfortunately, in many cases, the responses are below par. As a customer, I find it frustrating and lose trust in the solution when the AI-generated response suggests that I try the options detailed in my original query. I’m confident the AI responses will improve over time but until then, I, as the customer, need to hack the AI embellished systems.

I have to be careful not to lose energy over the responses and I need to figure out which button I need to push, or what words I need to say, to activate human support. Sometimes, I will put in something personal, like I’ll tell them a story about my dog, “Chester, my border collie, is hoping I can sort this out quickly so I can take him for a walk”, to help me figure out if I am speaking to an actual human. Who can resist talking about a cute dog?

The power of AI in forecasting and business analysis

Many financial analysis and forecasting solutions use AI-driven tools to help identify trends, visualise performance metrics, and generate actionable insights to support informed decisions. The customised reporting solution Etani has an AI analyst called Eva, who engages with you in natural language. You can chat to Eva 24/7 about your data, ask Eva questions, or request that Eva generate reports for you. Eva eliminates the need for skills in Power BI or even spreadsheets, which had previously been a barrier to using these powerful reporting tools.

Arcanum is building AI assistants to automate repetitive back-office tasks. Giving them human characteristics, Arcos is an AI customer support assistant, and Archie is your AI financial assistant, that you can connect with on LinkedIn.

Streamlining quality assurance and enhancing accuracy with AI

By its very nature, AI can identify anomalies which helps uncover errors, fraud or transactions that need a closer inspection across datasets. Addressing these issues swiftly improves informed decision making and operational efficiency and enhances trust among stakeholders.

The practice management software XBert connects your Xero, QuickBooks, MYOB or Employment Hero data files, and 88+ AI powered ‘berties’ scan your data, surfacing variances. The variances need to be reviewed and dealt with by a human. Other quality-check solutions include MindBridge and Dext Precision.

Maximising productivity with AI innovations

The workflow management solution Karbon has embedded AI, powered by Azure OpenAI Service to enhance email efficiency. It can summarise long email conversations and draft email responses, with the option to define neutral, conversational or direct tone. Karbon also includes the ability to summarise the last 10 emails and notes, helping you get up to speed with the customer relationship quickly.

The application Parabola combines AI with OCR (optical character recognition) and a vision-based LLM (large language model) to extract and contextualise the information from PDFs and transform it into usable data in spreadsheet format.

The workflow automation software Zapier features Zapier AI CoPilot, powered by OpenAI’s GPT-3. The conversational AI asks, “What would you like to automate?”, and turns ideas into Zapier workflows. For example, it can create invoices in QBO or Xero from rows in Google Sheets.

Additionally, Zapier can duct tape AI into workflows by connecting AI-empowered apps like Airtable, HubSpot or Zendesk. For example, it can create email replies with ChatGPT for new emails and save them as drafts so you, the human, can review them before they are sent out.

The writing tool Grammarly reviews your text and uses AI to give personalised guidance. You can choose whether you would like suggestions to ‘Improve it’, ‘Identify any gaps’ or ‘Increase the impact of your text’. I particularly like the option ‘Identify any gaps’ because it helps ensure I’ve provided enough information for the reader.

Cutting through AI hype

Many solutions have yet to truly have AI embedded; often it’s simply an application programming interface (API) call bringing in the large language models (LLM) magic. Marketing hype confuses us with the AI buzzword, a few thought-leadership blogs and ‘exclusive’ waitlists for AI features that have yet to be released.

We need to stay intellectually curious as this space evolves and use our judgement, critical thinking and communication skills to leverage AI and refine the output. Streamlined processes, increased efficiency and optimised operations through AI translate to massive productivity and profitability gains, and they’re within our reach.