AI Reshapes the Accounting Platform Wars

There was a time when the battle of the accounting platforms was like a fight between giant battleships at sea.

The protagonists bulked up and armed up, and the struggle was between who was bigger and more powerful.

Speed was nice to have, but not necessary, as it was all about the muscle.

API integration was the first game-changer, as specialized applications were plugged into platforms to perform very specific functions, adding value that, outside the big ERP systems, the mass of commercial software-as-a-service providers could never match.

Now, artificial intelligence adds another dimension as an evolved model emerges, enabling rigorous AI training to deliver new levels of domain knowledge and functionality to once esoteric, highly manual tasks.

To use a current, relevant analogy, the battlefield advantage is now with AI-controlled drones, while the legacy platforms are the targets.

This is an interesting side note to the so-called ‘SaasPocalypse.’ Yes, the SaaS accounting providers are under siege, but the clear path to survival lies first in integration, and then evolution.

So instead of being all things to all people, the big platforms can remain relevant — for the time being — if they collaborate with niche providers and serve as gateways to a suite of functions and applications.

Auditing pain

One of these niches is auditing, and we see this in the auditing of Self-Managed Superannuation Funds in Australia.

Australia has a globally recognized retirement savings system, with total assets under management of AUD4.5 trillion.

Around AUD1 trillion of those assets lie in Self-Managed Superannuation funds, where individuals and families can create their own funds and pursue their own investment strategies.

That is highly attractive to many people, but it comes with a compliance burden: annual auditing.

With over 660,000 self-managed funds in the Australian system, that is a lot of auditing that needs to be done, and it is a decent chunk of work for many mid-tier and smaller accounting firms.

For providers, this is a scale opportunity — as it is for accounting firms — but the product needs to deliver to funds that are often quite small, with only several hundred thousand dollars under management. At the other end of the scale, the largest SMSFs have over AUD100 million in assets.

Evolving to AI

One vendor offering an SMSF audit product is the Australian vendor Cloudoffis.

Cloudoffis has gone through several iterations of its solution, progressing from optical character recognition (OCR) to robotic process automation (RPA) to AI, deepening its domain knowledge and making it more dynamic.

“The challenge and pain point we saw was that even though every SMSF needs to be audited, there wasn’t much love in the technology space for the auditors,” says Manish Sheladia, one of the two co-founders of the company.

“Accountants were using technology tools, but they were more cloud-based templates, and I wouldn’t really call that auditing technology… The opportunity we saw was for a smart auditing platform which integrates with the accounting software, brings in and analyses the data and helps the auditor do the job in an automated way.”

The Cloudoffis AI trains on a library of observations, including common compliance issues for the Australian Taxation Office and automates, recognizes and categorizes a wide range of documents from every bank and financial institution a fund might interact with.

Before the human accountants even get the documents, AI has looked at guidelines, extracted information and flagged potential issues.

A third layer is moving into the agentic, which synthesizes the processing and presents reports back to the accountant for checking and interpretation.

All this is based on AI training that combines anonymized data and applies retrieval-augmented generation (RAG) to LLMs, vision learning systems (VLMs), and statistical models.

“This is our fifth generation of product, but we are moving to a sixth generation which uses the latest technology,” says Sheladia.

“We used to use a lot of RAG, for example, but now we have figured out that RAG is useful for some documents but not others, and we are getting much better at identifying that… Vision models are getting much better, and what was not possible before, because it was costly and slow, is now becoming more possible, so now we have a platform that allows us to ingest these new technologies very easily and create context and make it very specific for our customers.”

While much of the technology is new, the platform is not leaving behind trustees and accountants who rely on older tools. Excel, for example, is still supported.

“It could have been so tempting not to use Excel,” says Sheladia.

“But we realized that if you take Excel away from accountants, they will hate you for it, so we have kept it online.”

Back to the platform

While the Cloudoffis solution integrates with the established accounting platforms, the firm is not ruling out an evolution in which it might offer a single platform.

“The existing platforms reckon they are doing a great job, but there is a lot more that accountants do outside of what they offer,” says Sheladia.

Having created an auditing solution, the next step could be to break down functional silos and let smart AI tools work across a wider range of accounting functions, integrating them into a single platform.

This would be the next step in the evolution of the accounting platform wars, with AI the killer technology that might reshape the battlefield. 

Image credit: iStockphoto/AndreyPopov