–
Or listen using the podcast player below.
–
Transcript (Note: There may be typos due to automated transcription errors.)
Brian F. Tankersley, CPA.CITP, CGMA 00:00
Welcome to the Accounting Technology Lab hosted by CPA Practice Advisors. Your hosts, Randy Johnston and Brian Tankersley, are
Randy Johnston 00:10
Hello and welcome to the Accounting Technology Lab. I'm Randy Johnston with my co-host Brian Tankersley. There's a lot of new technology coming out this year. Today I want to talk about Materia, which has a website that's a little bit different. Try material.ai. This product is very well suited to the audit and assurance space. Our CEO Kevin Marlini and CTO Lucas Adams were both at Facebook before they came up with the idea for Materia. One thing to know about Materia is that it represents a trend of products where AI is built into the platform. AI will be part of the platform. I think a lot of people are worried that they're going to be left behind by AI. And our position is that you can have the general-purpose platform that we've talked about in other technology labs, chat GPT, Copilot 365. But for most accountants, we think AI will be part of the tools. It's all built in, just like we don't think of Secure Sockets Layer as being part of the web browsing experience. So, Brian from Materia, again, this is an AI platform for accounting and insurance operations, what do you think are the key takeaways for our listeners?
Brian F. Tankersley, CPA.CITP, CGMA 01:40
Yeah, I looked at their information and stuff. But the idea here is that it's a tool to analyze and summarize documents, summarize and write notes that explain the guidance, summarize the rules that are relevant. You can look at and review your engagements, you can look at and standardize your workflows. The key point here is, as I'm looking at this, if you have a lot of documents — leases and liabilities and stuff like that — if you have anything significant, I think there's a lot of value in that. You can use this to look at the documents that are relevant to you and understand and discover things. I actually read through and I saw that they have a privacy page. I was very impressed with what they said on their privacy page because they said your data will not be used for the following purposes: They have SOC 2, Type 2. They clearly state that your data is not yours. We do not train models or content that is provided to Materia. And they're saying all the right things when it comes to privacy. And I think that's important because, as you and I discussed, this is the AI world right now, it's a lawless place. And the free versions of all of their products, almost without exception, use your data to train their models. In fact, according to their privacy policy and QuickBooks Online and their website, they use your data to train the model, so I was very encouraged to see someone clearly say, “We don't use your data to train the model.”
Randy Johnston 03:44
Yeah. Brian, when you and I talked about this, these private AI models, even though they're private to your company, are a pretty significant advancement in AI. And Materia has said, we're US-centric. We try to work with your company's guidance and policies. We take in your client documents to help you with your day-to-day work. They read your company's checklists. Whether that's reviewing board minutes or debt agreements or whatever, they actually look for your engagement data to make sure everything in your working documents is connected. Recently, I think it was June 20th, Materia raised about $6.3 million in funding. This is important to know, because that particular round was led by Spark Capital, Haystack Ventures, Temps Thompson, Reuters Ventures, Exponential Founders Capital, and the Allen Institute for AI. They didn't disclose the amount, but I think we learned some amount along the way. But I thought the Thomson Reuters investment was significant. Because it gives us ideas that maybe this could be built into Thomson Reuters' Checkpoint Engage AI, that could be embedded more deeply into advanced flows, but the founders are trying to keep this relatively standalone for now.
Brian F. Tankersley, CPA.CITP, CGMA 05:37
Well, in general, as we look at this, I think what this is doing right now is very interesting. This appears to be a tool for large and mid-sized enterprises that are looking to use AI to process information. And one of the interesting things that they list there is where are their subprocessors? This is another thing that I look for on this page, where are they? They're very forthright in saying that they use Open A Eyes. So, on their Trust Portal, they list their subprocessors, and their subprocessors include Claude. They use Claude engine, merge, cohere, Amazon Web Services, Anthropic, and of course, Claude Data Dog is in there too. And they use AWS for hosting, and they're based in the US Eastern region. So, they're doing a lot of things. And they're working with multiple models. I think that's very important.
Randy Johnston 06:54
Well, Brian, I think that would be a great topic to cover in another tech lab, because you and I have talked about subprocessors. You addressed that at AI Confidential at K2 this year. And I concluded, Brian, that we should create a bad list and a good list of subprocessors, because we're concerned about what happens to our data when it gets to the subprocessors. But again, the subprocessors that we have here seem pretty good. Reputable companies, that's the way I think about it.
Brian F. Tankersley, CPA.CITP, CGMA 07:30
The only one I would hesitate to go with is Coherent, because I think they have a lot of connections to US intelligence.
Randy Johnston 07:40
Right. The sub-process and the whole conversation, if you don't know, we'll let you know in an upcoming Technology Lab. But Brian was right. Materia paid for a SOC 2 Type 2 audit, we integrate with Microsoft 365, various internal data and document stores, etc. Our implementation here is fast. We also offer pilots so you can evaluate the material as needed. And, as Brian pointed out, we're also very conservative with data. We give our customers access to their data, our customers' clients' data, so that our AI platform can answer questions or run workflows, but only the relevant parts of the data are sent to the AI model, and the data is never used to train the model. Now, in the previous Technology Lab, we talked about Open AI's chat, GPT. The $20 model is used to train the model, but the $25 model, which is the team's model, does not use the data to train the model. This is a change in policy for Open AI over the last year or so. But the material is intended to be used by all levels of the audit and accounting team, from staff and managers to partners to use to provide answers to clients quickly.
Brian F. Tankersley, CPA.CITP, CGMA 09:26
And I think, in general, that's the main story here. You know, what we're seeing again is that a lot of really, true innovation, moonshot-type projects are happening not inside the big companies or the big publishers themselves, but in startups like Materia. There are notable exceptions, and some of the things that we talked about, we'll be talking about at the next Technology Lab. Well, Thompson's co-advisory feature, Wolters Kluwer's answer, and the Connect tool are those two notable exceptions. But in general, you have to keep an eye on this space, because AI is going to be a hit, and everyone's talking about it. But it's not going to be long before some people are using it effectively. So I encourage you to do some experiments, look at some demos, check out some of these new tools to make your company more effective, more efficient, and more effective.
Randy Johnston 10:38
Very good. Brian, I think that's a good summary. The assistant in the material that runs the embedded assist, the embedded citation, the Microsoft 365 integration, the document workspace, it's all part of what this product does. Just to add a comment, the integration is not just with SharePoint and OneDrive, but with Box, Dropbox, Google Drive, and it integrates with a variety of other systems. Again, the situation is even worse. If you're in the audit business in the US, I think you should at least consider what effect the material has on your productivity. I look forward to talking to you again in an upcoming Technology Lab. Hello.
Brian F. Tankersley, CPA.CITP, CGMA 11:34
Thank you for your time. We'll be back next Saturday for a new episode of Technology Lab with CPA Practical Advisor. Have a great week!
=END=
