Deloitte develops audit technology for smaller firms

Register now

Big Four firm Deloitte plans to begin offering auditing technology next week to smaller accounting firms through a recently formed venture.

Deloitte created Auvenir as an in-house startup and tasked it with developing its own auditing technology that could be offered to small firms. Next week, Deloitte plans to announce the North American launch of the technology, known as the Auvenir Audit Smarter platform, which leverages artificial intelligence to help auditors with their work.

Auvenir doesn’t use Deloitte’s own proprietary auditing technology, but it was developed with input from Deloitte’s audit team. The Auvenir team also interviewed a number of small to midsized audit firms and their clients across North America to identify the issues they encounter with audits. It has been beta testing the Audit Smarter technology with several auditing firms in Canada.

Deloitte global audit and assurance innovation leader Chris Thatcher said he was tasked by his boss with coming up with the kind of technology that a startup might develop in competition with Deloitte. “One of the things he was quite concerned about, that kept him up at nights, was a couple of guys in a garage thinking about how you could do audit completely differently to how we would have done it traditionally in the past,” he said. “My boss basically challenged me to think about defending ourselves from disruption.”

His first hire for the new venture was Pete Myers, who became CEO of Auvenir. “We hired Pete to essentially build a team of startup guys, people from an entrepreneurial background and a technology background,” said Thatcher. “We did think quite long and hard about whether we should actually bring in some of the Deloitte auditors, but one of the things we realized early on was that at least in the early stages of the company we wanted to keep it autonomous, so we purposely did not bring in the Deloitte auditors other than from an advisory capacity just to make sure this could be ramped up just as quickly as it could be ramped up. But we’ve largely kept the business to operate as independently as we possibly could and create an independent venture which is now being launched to market as Auvenir.”

Myers set up the unit as a kind of skunkworks within Deloitte to develop the technology, and he ended up dealing with many other auditing firms besides Deloitte.

“From our perspective at Auvenir, we spoke to hundreds of auditors and clients, primarily in that part of the market, the smaller auditing firms and smaller clients, and what we found is that a lot of the technology that’s available to them is not the right size technology for the size of engagements they’re dealing with,” he said. “A lot of the technology is a little bit more complex and cumbersome for what they need. It’s really about building something that’s the right size for small engagements. Those small engagements could be taken on by a small accounting firm, or they could be small engagements that are being taken on by a larger firm. The other play in here is we want to open this up for any accounting firm. We’re not restricting who uses it. It’s available to any firm out there.”

Thatcher wanted Myers to take a fresh approach to the audit market. “What we challenged Pete’s team to look at was the problem of a new entrant to the audit industry, and we purposely didn’t ask them to think about what technology Deloitte could use,” he said. “We said we really want you to take a different view. If you were to enter the audit market today, what would you do as a startup, as a new entrant? Essentially what they’re doing is not just about technology. It’s also about creating a different type of business model, which is really creating a technology platform that audit firms will use, not just Deloitte, but all audit firms could use, and really trying to create a compelling platform for the industry, for the profession.”

The platform uses cloud-based storage, machine learning and artificial intelligence to improve workflow and collaboration between auditors and their clients.

“For the machine learning component, there’s three main parts to it,” said Myers. “Firstly, you need to have standardized data coming in, so we’re doing imports from accounting packages, and imports from banks. That allows the data to come in a standardized format and, more importantly, understand the format so we know what the data is there. The second part is having some ability for decision-making to go through on the platform, so we allow auditors to decide exactly what tests they want to run on that data and how they want to run those tests. We offer a lot of customization to enable that, and then we enable the populating of working papers for the output.”

The machine learning component helps auditors judge whether there is a low or high risk from the trends they are seeing. “It would be like saying nine out of 10 auditors when they were looking at this industry thought this was a high risk, but giving those insights in a real way,” said Myers. “It’s really providing those insights and tips, as someone who is working through the engagement or going through their decision-making to just have somewhat more confidence in the decisions they’re making. One thing that’s very clear, as a platform we’re not taking on the audit itself. We’re just a tool to be used by auditors, and we’re not making any of the decisions that need to be made in coming to an opinion.”

Deloitte itself would not be signing off on the audits. “The audits being run through this platform are not Deloitte audits,” said Myers. “They’re being signed off on by the auditor who is taking that on.”

Deloitte admitted last month to a data breach in which hackers were able to access client data from its internal email system (see Deloitte email platform and client data hit by cyberattack). However, Auvenir is using separate servers and is emphasizing security.

“We host the data, and it’s completely separate from where Deloitte hosts their data,” said Myers. “Data security is one of our most important governing principles. In conversations with auditors and clients, it’s top of mind that the data is secure. Basically all of our customer and application data is encrypted whenever it’s being transmitted between our servers or with a customer device. It’s also encrypted anytime it’s stored on our servers. The data is never stored or transferred to a customer without strong encryption and then our encryption algorithms are all compliant or exceed ISO and NIST standards. And in addition to that, we’re making sure we’re going through an independent SOC 2 certification to make sure we’re complying with the SOC 2 and ISO 27001 security standards.”

Deloitte sees the technology as an innovation for the audit profession. “We believe this is quite radical innovation,” said Thatcher. “This is radically different innovation from what we’ve seen traditionally from ourselves and from others in the market. This is the very first venture we have done as a global audit business and certainly we hope it is not our last.”

The Auvenir technology is also likely to be used by Deloitte to improve its own audit technology.

“We are working closely in a collaborative way with the Deloitte technology folks, and there’s a shared learning here,” said Myers. “While this is focused on an independent market and providing it to all accounting firms, the innovation also helps with Deloitte as well in strengthening Deloitte’s internal innovation for their existing clients. The Auvenir platform itself, at its very essence, is looking at helping the client and auditor have a better relationship together. This isn’t to say the relationship isn’t great at the moment, but we want to take away a lot of the mundane, painful tasks so they can have a better experience of the audit and focus more on providing insights and doing the areas of audit that they love, which is the relationship and insights and being a trusted advisor to their clients.”

For reprint and licensing requests for this article, click here.
Audits Audit preparation Artificial intelligence Machine learning Deloitte