MindBridge Ai Auditor in Ten Minutes
Join John Colthart, General Manager, Audit and Assurance & V.P. Product Management, as he jumps into the MindBridge Ai Auditor platform, demonstrating data ingestion, risk-ranked scoring and reporting, an overview of the Ai Auditor control point checks, flow analysis and workflows.
Good morning, everyone. I’m going to give you a really quick demo of the MindBridge Ai Auditor. So this is our cloud-based service. I’ve already logged in and what I’m going to do is I’m going to create a new engagement. So let’s say this is Audit 2019. Maybe this is a SAP user. Invites me to get started.
I’m invited to bring in some data. So I can connect directly to cloud-based accounting packages or I can pull in a data file that’s been taken from a finance system. So what I have is I actually have an Excel spreadsheet that I’m just going to drag and drop here. So what you see is it started to load the data file.
In a moment, it’s going to detect what kind of data file this is. So despite it being a spreadsheet, it’s actually determined that it’s a Sage 50 ledger. Prepares the data so it automatically loads it and then it starts the analysis. Now this analysis usually takes a couple of minutes. I’m going to actually speed up the video so we don’t have to sit and wait.
So as you can see, the file has completed the analysis. That was actually about a minute and a half of the lapse time. I’m going to click to see the analysis. So this entire dashboard has been created since we started this demo. So everything you’re looking at has been derived from that data file that we were just loading.
What we’ve done is we’ve looked at every single line of every journal entry and we’ve scored every transaction as a measure of its quantity of risk. So what we have is these high, medium, and low risk categories. You can see the number of transactions in each category and the dollar value associated. So it’s fairly easy to see that if I was randomly sampling transactions, I’m much more likely to find a low-risk one than a high or medium-risk one.
The idea here is that the auditor should spend their time looking at the most interesting transactions, the most unusual. So let’s scroll down, let’s have a look through this dashboard. I can see risk by day. I can also look at, sorry, risk by month. I can also look at it by day which gets me a kind of finer grained view. Let’s scroll along here.
You can see that there’s definitely a spike. You can see these are actually periodic spikes in this data. That’s fairly normal. When we look at accounting practices, you often see a flurry of activity around things like month-end closes. You can see there’s something definitely happening here and this is actually the year-end close cycle.
If I remove the low risk, they ought to see that far more clearly. There’s definitely a higher risk time period here. And I can click on any of these bars and drill into the detail of it. If I continue to look down the dashboard, you can see here on the right-hand side, here are all of the different control point checks that we’ve performed. If I’m interested in, let’s say, manual entries, I can click here and it will immediately take me there.
I can also look at some of the more detailed. There’s like our flow analysis, which is a machine learning based technique. On the left-hand side, you’ll see a visualization that we call the risk burst. So previously, we were looking at risk by time or by risk level. This is actually breaking it down by accounting area. So the concentric rings here, the innermost ring is the highest level of the chart of accounts.
So that’s your assets, liabilities, revenue, and expense. And as I move out through these concentric rings, I’m going out, further down into the chart of accounts, into more levels of detail. So you can see here, if I go out here, I’m getting to one of the more detailed accounting areas. The system definitely thinks that there’s something going on in the equipment transactions. And I can just click on that slice.
It takes me to a ranked list of transactions that impact the equipment account. And I can look at any one of these and drill into it in more detail. So let’s pick this one. So I’ve gone, I’ve drilled all the way from the highest level looking at the overview of the ledger as a whole, the business as a whole and I’m now looking at a single individual journal entry.
So you can see it’s received a 35% score in terms of overall risk and what you’re looking at here, as I scroll down, you’ll see all of the different control point checks that have been performed. The ones that are green, the system has said, “No, it doesn’t fail this test.” But the ones that are red are the ones that have definitely highlighted this one. So to explain how our AI had selected this transaction, all you have to do is read off from these kind of red items.
So this is a transaction that happened at the end of the year. It was a high monetary value. It was entered manually. But more interestingly, it’s a rare financial flow. That’s a machine learning technique where our AI is looking at all of the different account interactions and it’s determining that this particular accountant interaction is unusual.
It’s also an unusual amount, which is the next machine learning technique. So a combination of these different control points is expressing an opinion on this transaction. They’re all combined together into the risk-based score and that is what has brought us to this transaction being interesting.
If the auditor wishes to follow up on this particular item, they think that this is indeed unusual, they can create a task and I can assign it to one of the colleagues working with me on this audit and I can put an alert in and this creates a task which is a part of our workflow. In the system, I can go to my audit plan, review the tasks that are currently open.
And I can even go to the reporting area and I can get…I can produce reports on different aspects of the auditor, different parts of the business including things like key performance indicators. So that’s your five-second tour or a two-minute tour of the MindBridge Ai Auditor. I hope you enjoyed it.
CPA roundtable: Planning an AI roadmap for 2019
Join our industry panel of CPAs that have gone through the diligence and deployment of adopting AI-based solutions for their firms and offer advice on how other firms can get ahead of the technology game.
Coffee with a professional in accounting: Samantha Bowling
Join Samantha Bowling, Audit Partner at Garbelman Winslow (GWCPAs) as she chats with us about her strategy for success and innovation in the audit industry and how her firm took the leap into AI with our MindBridge Ai Auditor.
3 steps towards AI-enhanced audit by 2020
Artificial intelligence isn’t a dream – over 40 of the US Top 100 are using MindBridge Ai Auditor right now. You’ve heard the news and hype, but do you know how to make AI-enhanced audit a reality for your firm?