Based on headline-grabbing generative AI breakthroughs such as OpenAI’s ChatGPT, this new AI handles administrative and repetitive tasks in ‘real time’, whilst recommending ways to make savings and drive business improvements. It helps with forecasting, cashflow management and generating and sending invoices, all via simple, natural language commands.

In short, it really is like having a human assistant. You talk to it. It talks back. It spots things you need to know, and it tells you. It really is a game changer.

Sage got the march on Xero with Copilot, its AI assistant that’s already available for early access.

The Power of Data

Regardless of the ledger you use, what powers the new AI revolution is accounting data. This is the secret sauce. After all, the AI can’t make helpful suggestions about your cash flow if there’s still a stash of paper receipts in a folder, waiting to be entered.

Just as with any human, the AI simply won’t be able to get the insight it needs without the right information. And this new AI really does dig down deep into the data, and also understands what it is. That’s why it’s so powerful.

What’s required to prepare now for AI assistant tools like Copilot are two things.

First, the accounting data flow needs to be always-on, so that it’s effectively 24/7.

Second, the data needs to be of a high quality. The general rule of thumb is that, if you feed bad and inaccurate data into a system, then you’ll get bad and inaccurate information out of it.

Luckily, there’s been a lot of work over recent years to try and ensure there’s a fast flow of high-quality accounting data.

For example, here in the UK we’ve seen open banking introduced, that makes it easy for accounting ledgers to hook into bank feeds so that they can get live data about transactions.

This takes care of accounting data that’s already in the digital realm. But what about all that paperwork businesses receive on a daily basis, like invoices, bills, receipts and statements?

Okay, so some suppliers have tried to make life easier by emailing PDF versions of these documents. But there still needs to be a way of getting the data out of the PDF and into the accounting ledger.

Data entry

This is where AutoEntry comes in. You use your phone to take a snapshot of paperwork when it arrives, and we extract the data. You can then categorise it, and click to publish it straight through to the accounting software.

AutoEntry will even try to autocomplete the categorisation, based on what you used previously. You can then create rules so that the entire process is automated. Just take a snapshot of the paperwork and we’ll do the rest.

Many people use their phones to take a snapshot of the paperwork—and if you can take a selfie then you can use our app. But you can also use a desktop scanner. This is useful if there’s a lot of paperwork to get through, such as if a client drops off their entire year’s accounting on your desk on 31 January.

It’s the same with those PDFs you receive by email. Just forward them to your special AutoEntry email address and we’ll extract the data in the exact same way. You can even provide the supplier with your special AutoEntry email, so the PDFs are emailed straight into AutoEntry.

Automation

The new AI breakthroughs are powered by something called machine learning. We know all about this at AutoEntry, because we’ve been perfecting it within our data automation extraction tool for years.

It powers the automation that makes AutoEntry so invaluable. And it’s automation that you’ll need to focus on to prepare yourself and your clients ahead of when the AI assistants go mainstream.

What can be automated in your working processes? You need to think hard about it.

For example, did you know that there are apps for your phone that can measure car mileage? Examples include MileIQ and TripLog. They not only automatically track your mileage, but they do so in a way that’s compatible with the government’s requirements.

Ultimately, it means no more manually jotting down mileage figures, or working out the calculations. The data is right there, ready and waiting for you. TripLog sends the data through as an invoice, for example.

It’s yet another data flow that you can automate. And that’s what we need more of, moving forward.

Digital tax

Even if you’ve no plans to adopt AI assistants (which would be a mistake—you should at least take a look at them), there’s another reason why you should be focussing on 24/7 fast flows of high-quality data: the government-mandated digitisation of taxes.

Here in the UK, we have the Making Tax Digital initiative, which already applies to VAT. Next year we’ll all need to start preparing for the introduction of Making Tax Digital for Income Tax (MTD IT), too.

Accountants and their clients need to get into a whole different frame of mind when thinking about their accounting data. Yearly check-ins just for compliance are no longer an option. Data needs to be 24/7, and accurate.

It’s no mistake that the big accounting software companies like Sage are aiming to have their AI assistants ready for the introduction of MTD IT. AI assistants are going to make complying with MTD IT’s demands significantly easier for everybody.

Why AutoEntry?

When it comes to adopting the new AI, we need to focus on getting that flow of high-quality data.

AutoEntry is designed from the ground-up for just that task. Accuracy is up to 99%, including human review of any data our computers are unsure about. (With industry-standard data privacy protections, of course.)

With AutoEntry you pay only for the data extraction you use, via a monthly subscription that you can cancel at any time (up to 24 hours before renewal). Everything else is unlimited, including the number of client companies you add, the number of user accounts, and even the amount of cloud storage you use for the documents.

Why not sign-up for a free trial of AutoEntry? It takes minutes. There’s no need to sign-up to a contract, and you’ll get 25 free credits. You don’t even need to provide any payment details, such as a credit card.