Windward Insights

Leveraging Artificial Intelligence in the Financial Services Industry

Published
Written by Sarah Wolfe

video screencap of Matthew Dean

Insights from an interview with Sean Mcdermott and Matthew Dean from Freddie Mac on the AIOps Evolution Podcast

Matthew Dean has been in the financial services industry for 15 years and is currently at Freddie Mac working in the single family data and decisions department. Dean sat down with Sean McDermott, the CEO of Windward Consulting to discuss the role of artificial intelligence in IT Operations (AIOps) across the loan processing lifecycle. McDermott and Dean discuss the importance of the operations side of technology and automation, the future of IT operations, and how to monetize data.

Machine learning reduces appraisals in loan processing

Businesses continue to be driven by technology and automation. Freddie Mac has leveraged automation and machine learning to reduce the need for appraisals and to create more efficiency throughout the loan process. By leveraging a set of standards for low risk loans, Freddie Mac has been able to reduce man-hours and provide even more value to their end customers.

Ultimately, their goal is to make homeownership accessible and affordable. Dean demonstrates how automation and machine learning can deliver on that objective for Freddie Mac in several use cases. The three most important areas he focuses on are:

  • Proactive customer engagement
  • Increasing operational efficiencies
  • Minimizing risk for the business

The Future of IT Operations

Moving forward, it is essential that automation and machine-learning are part of your IT Operations and AIOps strategy to keep up with your competition. McDermott shares that there are three core factors for moving the business forward; mission, people and technology. Freddie Mac is a great example of how to combine these three factors into a strong IT Operations strategy that moves the business forward.

Monetizing data

The data economy is here to stay and smart IT teams are learning how they can monetize their data and become part of the revenue equation for their organization. However, it doesn’t come without risks that need to be managed. Customer data privacy is an incredibly important consideration when looking at data monetization.

It can also create an opportunity when viewed as an asset of the organization. Finding the balance between privacy and opportunity is an art and science that requires IT to partner with the business and legal to uncover where it makes sense and where it doesn’t. Dean shares his thoughts on the data economy, and how he is moving Freddie Mac into the future of data intelligence.

Dean and Mcdermott provide a fascinating discussion on the importance of automation for loan processing. As IT Operations evolves and progresses, organizations must be ready to adapt to these changes in order to reap the benefits of new advances in technology.

Want to dive deeper? Watch the full episode below, listen here, or view the transcript below to catch the highlights.

Show Notes:

Matthew Dean has been in the financial services industry for 15 years and is currently at Freddie Mac working in the single family data and decisions department.

https://www.linkedin.com/in/matthew-dean-9130b034/

http://www.freddiemac.com/

Transcript:

Speaker 1 (00:01):

Welcome to the AIOps evolution podcast. This series features visionary IT leaders who are paving the way for the next evolution of the IT industry. Discover the truth about where we are in the adoption of artificial intelligence for IT operations and actionable tips that you can implement to become a more effective IT change agent. In this episode, we are joined by Matthew Dean. Matthew has been in the financial services industry for almost 15 years. Matthew’s early career began as a trader for Trillium trading, and later as an analyst for Morgan Stanley. For the last four years, he has worked with Freddie Mac single-family data and decisions department. His specific roles have included building and leading a team focused on business monitoring and machine learning for the purpose of developing deep insights and loan efficiency through micro and macro process management. His passion is to help turn data into something meaningful and to provide clarity into business processes. Now let’s join your host, Sean McDermott, a mission-driven serial entrepreneur, IT engineer, and AIOps visionary for this exciting discussion. Welcome, everybody.

Speaker 2 (01:14):

The next installment of the AIOps evolution podcast with me today is Matthew Dean with Freddie Mac. So welcome Matt to the podcast.

Speaker 3 (01:25):

Oh, thank you for having me on to talk. Yeah, yeah, you got it. You nailed it there.

Speaker 2 (01:35):

So I’m excited to have you on because we talked to a lot of vendors, we talked to analysts, you know, so we always like to talk to customers, right. And what’s going on, what you guys are doing and what you’re, why you’re implementing AIOps, and what you’re trying to get out of it. So let’s, let’s start by understanding your area of expertise. So I’m really actually interested in your background as a trader, right. And now in the IT department, you know, working on data analytics. So how did you make that transition from being a trader, you know, in New York City to…

Speaker 3 (02:14):

Yeah. Sure, sure. So, great question there. So coming out of school now, I was looking for something exciting. I was into finance. I went to school up in New York, so everything centered around finance. So I went in, became an equities trader. So it was some exciting job. It’s a lot of what you see in the movies is happening on the trading floor. So it was a fun time, but what I found with trading was got to learn about the equities market pretty deeply there. However, there are a lot of skills that I saw, you know, I wasn’t picking up while working as a trader and I kind of saw this resonate when I was at some of the older trailers. And when I say older, I mean above 30 years old at the time here, so they might’ve had pretty good careers making large sums of money, but they hit 30 and, you know, maybe they weren’t trading as well.

Speaker 3 (03:14):

It’s a little bit harder to take on risk, the older you get, you get more of a family, and yeah. So I mean, I’ve passed that by a decent amount now, but at the time it was old. So I wanted to go and you know, and get a little bit more into the business side in terms of a bank and its operations there. And then that naturally transitioned, you know, as a lot of what we think of, of business functions are being driven more and more by technology and automation and naturally led into coming into the more focused on it side there. So that’s kind of how I made the transition. Then 2013 moved down to the DC area there. I was still working at Morgan Stanley, but at one of their satellite offices, so kind of a little bit different feel than working at the headquarters. So transitioned to consulting for Freddie Mac for a few years. And they asked me to come on full time and I jumped at the opportunity.

Speaker 2 (04:32):

So yeah. So it was all your expertise, right. And making these transitions. I think you bring a very interesting dynamic to the conversation, right. Because a lot of it, people start in it, right. They go to engineering and then they kind of rise up through the ranks. You’re going to start as a system administrator, but from your perspective, what is your, what do you, what do you perceive as the potential for business monitoring over the next several years?

Speaker 3 (05:02):

Yeah. So great question here. And I do think that I have a very business-focused view on the technology side here. And my view is that you know, I teach here to support the core functions of a business. And as business processes are driven more and more by technology by automation. It’s very important to ensure that the business side has insights into what is actually happening here. And it’s for a few reasons, the most important in my opinion is that they can address any customer concerns here. You know you don’t want to get into situations where a customer tries to use your system, Freddie Mac, let’s say they try to sell us alone. They can’t do it. You want to be able to proactively be able to see it like, “oh, bank of America, wasn’t able to submit loan XYZ”.

Speaker 3 (06:08):

Let me give them a call and see what’s going on there. You know, I think that sometimes on the technology side, where there’s great technical monitoring, but it might say like a server’s down. Well, so what does that mean? That transactions are failing? Is it affecting our customers? So I think the business monitoring gives great insight into addressing those customer’s concerns next. I think it reduces risk if your business people don’t have a great insight into what’s happening in the systems that drive their processes, it’s going to be hard for them to manage risk grieving, know the level of risk they’re undertaking when they do, you know, sign off on a deployment of some new upgrade to a system here. So I think it helps them manage that risk. And then it’s gonna be better for the efficiency of operations here.

Speaker 3 (07:06):

For example, they’ll be able to view how many transactions are going through a system and find out exactly what failures happen for which customers, without having to go to it, to go and ask for a report here, or, you know, maybe they are a little more tech-savvy, dig into the system themselves and create that report business monitoring provides that for them. So I think that those three areas there, the proactive customer engagement increasing operational efficiencies, and reducing risk, I think business monitoring helps tremendously there. And, you know, I think it provides a lot of value.

Speaker 1 (07:53):

We’d like to take this opportunity to tell you a little bit more about our sponsor Windward consulting group for over 20 years, Windward has been helping IT leaders run complex global networks and data centers to deliver impactful business services. Windward is the full package with industry-leading solutions, architects, and integrators with certifications, partnerships, and deep expertise to integrate leading ITS and AIOps tools with leading technologies like Service now and Splunk Inc. From technology implementation to integration to managed services, Windward has your needs covered with Windward you get strategic thinking and real-world results. Do you want to see if Windward’s ATM is a good fit for your needs? Go to www.windward.com/AIOpspodcast to book your discovery meeting today.

Speaker 3 (08:49):

So what do you see AI for all and machine learning? And business monitoring going forward? I think of AI helping out in two main ways here. First and foremost, I think it’s going to help us drive business decisions here and we already are doing that at Freddie Mac. For example, we have a program automated collateral evaluation, where for certain loans, if they meet certain criteria, we are not requiring a fiscal appraisal of the property. This is great all around for the borrower, you know, and we want to support the borrower to our mission and home possible. Um, it means that the same money on closing costs, they don’t have all the anxiety of going through the appraisal process. It reduces time to get to closing so great for them there. It’s great for the seller. They have more purchase certainty. They go through the program, they’re able to see that, Hey, we’re going to buy it this alone, even without the appraisal.

Speaker 3 (10:00):

So in their pipeline, they’re not going to get a kickback at the end here. And for us, we have better data on the loan. So it helps us manage our risk more. That’s driven through, you know, all the data that we’re collecting through AI here versus having a single appraiser go out there. You know, they’re having a bad day. Maybe they take a few pictures and look at one or two cops. Don’t do a great job. We have the data to make sure that we know the loan coming in is good and helps the borrower out. So that’s just an example of how we’re using data to improve some of the business decisions here. And then I think it’s going to help us and our operations, for example, we can use AI to look at patterns of when we can anticipate failures are going to happen.

Speaker 3 (10:53):

So let’s say we see CPU utilization kicks up for three days in a row on a specific server on the fourth day through an algorithm, we say, well, there’s a 50% chance that that’s going to cause failures within the system. Well, now we can even get automation in here. We can say, if we see CPU utilization tick up for three days in a row at night, automatically run a script to bounce the server. So this saves, first of all, it means we can reduce failures proactively. That’s great. All around helps customers help us. Secondly, it helps with our people here that they don’t have when there are failures, they don’t have to troubleshoot it, stop it before it happens. Secondly, our IT professionals are doing higher-value tasks. They’re not going in and doing something as mundane as bouncing a server. So I think that those are the two ways that I see AI helping a lot, you know, helping with our business decisions. I think that that’s number one, but also helping with our operations.

Speaker 2 (12:16):

Sure. So is this, cause I was going to ask you to give us an example of where you think AI can help eliminate, and you did that. So I’ll shift a little bit, the question, one of the things we talk about a lot on this podcast too, is the need for AI, right? And really, it’s not about reducing headcount. It’s about reducing the gap between the ability of your team to process data. When these IT systems are kicking off more and more data all the time, and that data gap is getting much broader. Are you seeing that at Freddie Mac, are you seeing the amount of data being processed, you know, going up regularly?

Speaker 3 (13:00):

We definitely are seeing that you know, we rolled out this new program. Well, I guess not new, you know, maybe about half a decade old or so loan advisor suite here where it’s great. We evaluate loans before they come into Freddie Mac to help with purchase certainty, making sure that we’ll buy them. If they meet certain criteria, we can get rid of reps and warrants there, but there’s a lot of data that’s coming in with that. You know, we want our in, you know, with all this new data and the new systems that are around to run it here. Um, we want people, you know, first step to insights into those systems there. And secondly of doing them more high-value tasks of saying, thinking about, well, how are we actually going to put these together?

Speaker 3 (14:06):

Um, so having AI to help with our operations, like in the example above will mean that, you know, our developers won’t be troubleshooting issues. They’ll be helping to build out new systems and it’ll mean that our business people, you know, in terms of the business decisions that we can help automate like an ACE program, we’ll be able to think of new ways that we can help our customers. They are, instead of going through, you know, and manually evaluating does a loan need an appraisal, a fiscal appraisal, or can we use data that we already have? So yeah, hopefully, that answers your question there.

Speaker 2 (14:50):

I, you know, we talk a lot about the amount of data coming in and just want to make sure that we’re, we’re talking correctly and because we hear that from a lot of customers. It’s good to hear it from your side. So you know, kind of leading on to a little bit deeper into the applicability of AI and helping reduce mundane tasks, do you, how do you guys think of this from using AI from a cost savings perspective? And I know we talked about time savings, but any other thoughts on that, but how do you think about AI from an ROI perspective, and cost savings?

Speaker 3 (15:32):

Yeah, so I, I think that in terms of cost-saving here and I think with business monitoring, you know, that’s, what I’m focused on here is being able to proactively see what problems are going to occur and address them beforehand. I think that that’s a huge cost savings. Um, you know, if there are failures, we have to spend time, you know, and I know that you said, well, think about some other items, but I think times, you know, that’s what we spend a lot of our money on paying for resources. And we want our resources to be doing high-value work, not troubleshooting issues that are happening within our systems here. Um, so I think investing in ways that we can proactively stop problems before they occur, we free up our smart people to do things, you know, that the computer can’t do more of the critical thinking.

Speaker 3 (16:36):

So that’s one way. And the other thing, like I go back to the business decisions here, more data, you have more information, you have making decisions. If we can go into our risk models, you know, that has a direct impact that we’re making sure to not take on too much or too little risk. Um, that’ll help our bottom line and our underwriting models here, making sure that the loans that we bring on we’re pricing them correctly. Now that’s our bread and butter and we’re, we only will get better with the more data that we have if we price the loans correctly, that means more revenue

Speaker 4 (17:24):

Introducing red monocle, the only commercially available software solution that provides IT executives with real-time, AI-driven analysis of their IT investments and associated risk factors, Red Monocle simplifies tools, discovery tools, rationalization, and risk analysis. So you can achieve always-on service availability, mitigate the security risks of technology acceleration and optimize it. Red Monica is a single source of record for it. Tools, costs, and features powered by a knowledge base of over 1500 commercially available tools run. What if scenarios to see the full impact of technology change and identify gaps that may result in security breaches and downtime, and you’ll get an expert portfolio analysis from your dedicated portfolio advisor, red Monaco is a CIO, his best friend, and turns your IT tool portfolio into a strategic advantage. Want to take a test drive, go to redmonocle.com/AIOpspodcast to book your demo.

Speaker 2 (18:25):

Let’s talk a little bit about the applicability of AI in today’s economy. And how do you see, let’s kind of shift off AI a little bit and let’s just talk about operations, right? I’m an operations guy. I’ve spent my entire career working with companies in IT operations. Where do you see IT operations going over the next five years or 10, that’s fine. And where do you think that IT operations teams need to transform? And one of the things that I’ve talked a lot about in the past is a lot of people perceive operations as a sunk cost, right. A cost of doing business. And I think if, if organizations could think of operations as a competitive differentiator, right. Does that type of thinking go on where you guys are at? What do you think about where operations are going over the next few years and how to really make it more of a competitive differentiator in the market and what you guys are doing?

Speaker 3 (19:41):

Yeah. So I guess in terms of, you know, I’ll take on it as one and operations and as another, we’ll see how this goes. But I view these functions as supporting a company’s core competencies here. And for some companies that is operations, you know like UPS, FedEx, they’re supply chain management, they’re in their operations. That is their core differentiator here. I view it in operations, they’re supporting these core competencies and we can look at Walmart example, they’re spending a lot of money to be fought there in IT, so they can be an online realty. Our retailer competes with Amazon, you know, brick and mortar stores are kind of going out, but their core competencies are having massive buying in their supply chain management, even with all the money that they’re investing in.

Speaker 3 (20:52):

It’s to support the advantages that they have in the marketplace for Freddie Mac, you know, our core competencies are to buy securities and manage loans and the risk associated with it. That’s what our it and operations aim to do. Now, the better IT operations, you have, you know, the better customer engagement, you know, you get, they can answer their questions quicker. You can give them easier platforms to interact with the companies. So, but it still goes back to the core competencies for Freddie Mac is the support, rest buying loans, and securitizing other companies who have core competencies around it, like Amazon, you know, like they’re taking advantage of it, they’re coming out of AWS there, and the cloud and that’s what the company’s advantages, you know, they should go with it there. But I think for other companies you know, Freddie Mac particularly focused on that, that’s where I work. Um, you know, it’s to support our overall mission here of getting the loans you know, and in our mission statement, making home possible. I think so, I think so. I don’t know. Maybe that’s a little bit different than what it sounds like you might have a little different view there.

Speaker 2 (22:24):

No, it’s a, it’s truly no wrong answer. Right. I think, you know, when I look at, you know, when I look at successfully running companies, right I take a little bit of a different spin on what it takes to run a successful company. And, and it, to me, I look at three primary areas. One is the mission of the company. The second is the people, and now I actually shift to technology, right? And maybe I’m a little biased because I’m a technologist. And I know that some investors talk about product-market management, right. Mission to me is part of that product. What are you trying to bring to market? And if people are people, but to me, I think technology is now becoming the staple and any company, I don’t care if you’re at a restaurant, you know, or your, you know, Walmart or Amazon or Freddie Mac investing in technology.

Speaker 2 (23:25):

And this is why it is here to stay. Right. And even more, because it’s becoming a huge differentiator. And I think your, your, your example of no collateral, right, that program is really a really good example of leveraging technology to gather data from all these different sources, to be able to process loans faster and cheaper and not rely on the manual, the old manual process of somebody walking out there and like going out and taking pictures and getting around to it whenever they feel like it and things like that. So, you know, to me, I think that what’s exciting about technology is that it’s becoming, it’s big, really big. It has to be part of everyone’s business strategy going forward. And certainly, we do work with Freddie Mac, right? So we’re, we’re worth you guys. And we see that every single day and have how Freddie is, is investing in technologies to, to, to really support that core business and expand services, make their core business better, but it was also expanded in other areas and leverage that technology.

Speaker 2 (24:29):

So it’s interesting to watch what you guys are doing over there because I’ve been in the DC area for 30 years, right. I’ve known one of my very first clients, actually, my very, very first client when I started my very first company was Sallie Mae student loan office. Right. And building monitoring systems to monitor FTP files coming in, from people originating student loans. And they would FTP a file. And we would, you know, I built a monitoring system to track it all the way through to the mainframe that would then process, you know, process the file. So it’s, it’s just been interesting to see how much technology has really moved the needle on how everything has done today and where it’s going to go. Yeah. So good. So so I know you’re passionate about data and this is an area many companies do not recognize as an asset, they call it. So in this economy, right. What are the opportunities you think there are for companies in capitalizing and monetizing their data?

Speaker 3 (25:38):

Yeah, so I think data is absolutely huge. I mean, it’s like oily, it makes everything go. And I think really I view it like a lot of things, two big ways that people can use their data and monetize it here. And I think the more practical one and the one that companies should strive for is to use it internally to inform their business decisions here, you know, like how can risk models be strengthened with the more information that we have? You know, we don’t have to have question marks in there. We have fewer question marks. Um, how can we strengthen our underwriting models? And we can use all this data that we’re gathering to do that. So it’s to inform business decisions. Um, in other ways, internally it’s improved efficiencies with the operations, you know eliminating the need for that appraisal, you know, I keep going back to it, but I think it’s a great program.

Speaker 3 (26:55):

You know, that’s one step that we can eliminate because we’re using the data at hand and it’s a win-win um, when, you know, for us sellers in the borrower, it’s, it’s great. Um, companies can use it to enhance their client experience. You know, Freddie Mac can be seen as reaching out to customers proactively now where we see that, you know, usually you’re submitting a thousand loans a day. We have nothing for you today, something going wrong, you know, people appreciate that. Customers appreciate that. Um, you’re looking out for them. Um, you know, and I think once again, I’ll get back to our mission, all these trials, all these ways that data can help Freddie Mac and the examples I give, go to the underlying mission of making home possible and, you know, making sure that we can have stable secondary market for mortgages here.

Speaker 3 (28:00):

So that’s one way, and I think that’s the more practical way that companies can use their data, the other way that data can be monetized, it’s pretty obvious here is it can be sold, you know, for example, Facebook here selling it to advertisers. Um, so that’s very lucrative, but I think that you know, there’s a lot of privacy concerns around doing that. Um, I think that the world as a whole is still figuring that out. You know, there are different laws in Europe now. I think if you get a picture taken of you, you own your likeness in that picture. Um, no, but there are different laws in Europe than there are in America. We’re still figuring out the privacy concerns. Um, I believe, you know, people’s privacy is very important. I think Freddie Mac has that same view. We have a whole entire privacy office here. Um, that sets out policies that the firm follows in my area, you know, data and decisions. It’s, you know, pretty technology-focused, but tactically, we do things like ensure that we heavily restrict access to any KPI data on a need-to-know basis. And that’s that, you know, sign up DNA, our NDAs to get access to some of the systems

Speaker 2 (29:36):

As a partner. You guys, we had to, we, we sign all kinds of agreements with you guys on PKI. You guys definitely take it seriously.

Speaker 3 (29:43):

Oh yeah. Yeah. I am. It’s important. You know, the borrowers they’re giving us information, we have to be stewards of that. And we have to make sure that we’re protecting that. And I think that we do there, you know, we’re masking data and systems we’re taking precautions there. So anyway, I think that those are the two ways, I think, like everything we discussed earlier here in a way that companies can use it internally, you know, I think it’s the few examples I just gave touch on it. A lot of points that were discussed earlier, then there is definitely the side that you can sell this data. Um, companies can do that. Uh, it seems risky there. We have to get that sorted out the privacy concerns, but people do and make money doing it.

Speaker 2 (30:37):

Yeah. Well, that’s a whole other conversation on the business.

Speaker 3 (30:43):

Um, all right.

Speaker 2 (30:45):

Well, I appreciate your time. It’s been a pleasure talking to you. I really appreciate you coming on this podcast and talking about it. It’s always, always interesting to hear from customers and what you’re doing. So any parting thoughts?

Speaker 3 (31:00):

No, no, thank you. It’s been wonderful working with your team and there, and I really appreciate you having me on to share my thoughts.

Speaker 2 (31:11):

Awesome. Well, thank you very much. And I’ll, I’ll give you back some time in your day and get back to the real important work for you.

Speaker 1 (31:20):

Perfect. Thank you. Bye Matt. Thank you for joining us in paving the way for the evolution of AIOps and the next generation of it. Innovation. If you are inspired to dive deeper into the movement, check out all the resources available@aiopsevolution.com.