Immerse yourself in an engaging series of technology-focused presentations, unveiling an array of innovative products and services specifically tailored to enhance your workflows.
Transcription:
Panel Member 1 (00:10):
To kick us off today, we have Suketu Gaglani and Bhagavath Bodagala at Mira Labs. Go ahead and kick us off.
Suketu Gaglani (00:22):
Thank you. Good morning everyone. My name is Suketu Gaglani and I'm the Founder and CEO at Mira Labs Ai. Over here we have Mr. Bhagavath Bodagala. He's the Head of Engineering at Mira Labs Ai, and today we are excited to introduce you to a state-of-the-art loan origination system point of sale integrated with AI called Mira. Mira is an AI copilot and in our traditional lending process I have worked with in my entire 18 years of journey, I worked with many mortgage lenders, brokers, independent loan officers, IMBs, and all of them face a similar challenge in achieving their KPIs. Does this screen look familiar to you? Look at the numbers. These are our KPIs that we always struggle to achieve. It can be from multiple factors. It can be from market volatility to uncertainty to inflexible operating hours, to limited access to tools and resources. All of these factors are impacting, our loan officers are limiting our loan officers from focusing the most important thing, which is client engagement, sales growth, and high customer satisfaction.
(01:39):
On the other side, we have our operations and processing. We all know how highly manual with cumbersome workflows our operations are right now, and all of these are adding to our inefficiencies and unplanned capacities. And then all of these things together are impacting our borrowers who are the ones getting most impacted with their long wait times, lack of transparency, and a very unpleasant mortgage experience. To address all these three challenges we developed, Mira, let's quickly look into the demo and see how Mira can transform your journey and enhance your efficiency and experience throughout the mortgage journey. That's here to believe it.
(02:25):
As I told you, the Mira is a point of sale and a loan origination system. I have a borrower who's about to start his mortgage journey and has some queries he wants to get answered before he starts his market journey. At this point, they can easily access the point of sale provided by their loan officer and they can use MIRA to get all those answered without waiting for the loan officer to call them and address those issues. So let's say refresh. Okay, I'm so sorry. Yeah, yeah. I need to log out and log in. No, don't do it. Refresh it.
(03:03):
So sorry about that. As Mira is an AI copilot, it has the capacity and ability to go ahead and access all the guidelines that has been embedded into Mira to go ahead and get all this thing done immediately. It can provide you a very natural language understanding of the entire process in a very easy to understand for the borrowers, which I would've done to my borrowers today. So let's say I've just gone ahead and asked a query that I'm putting less than 20% down and why do I need to pay PMI MI will explain that in a very easy to understanding language along with providing those reference links as well. Let's take a step forward once I've got all my questions answered, what happens next? Many times I've come across borrowers who have gone ahead and asked for eligibility just to understand based off their income, how much do they really qualify for. I don't want any credit pulled. I don't want any documents, but can you just let me know how much am I qualified for? Well, now Mira can do that for you. As you can see, Mira is going to be able to go ahead and pull out all the details from the profile. Let's go to the profile.
Bhagavath Bodagala (04:12):
Yeah,
Suketu Gaglani (04:14):
Yeah. So it can already pull out all the information embedded in the profile and can go ahead and run your eligibility checks as well. Let's look at the next step of where a processor is currently reviewing all the documents manually, I have gone ahead and provided MI with a document of a pay stub, and I'm asking the Mira to go ahead and review the documents for understanding the income consistency. Mira's AI document intelligence can not only read the document, understand the document, review it and provide you all the questions answered immediately. I'm so sorry, just taking a little time, but it's ai.
Bhagavath Bodagala (04:58):
Yeah, Wi-Fi is little slow. Yeah, that's okay.
Suketu Gaglani (05:01):
Yeah. While this is happening. So Mira is built with AI document intelligence that can actually read your documents, extract the data of your documents, and go ahead and let you know what exactly you want to know based on those particular documents. So let's say I have a question that this is a pay of Mr. Suketu. Can you please explain the consistency of the income? Is it right in line or no? Mira can extract the data from this and give you a very easy to understand calculation as well. Let's look at the loan officer scenario. I've been a loan officer myself for 18 years and for 18 years I've come across so many scenarios in my life that whenever there's a difficult scenario, I have to dig into tons of guidelines or probably reach out to an underwriter to get an understanding whether or not a loan is doable or not.
(05:57):
In today's time, we have just eliminated that step a lot. So the moment, let's say I have a scenario over here where a borrower has multiple credit derogatory accounts along with the foreclosure that's been reported in the past. So for me, I have to go ahead and I have to go ahead and dig through all the Fannie Mae, Freddie Mac guidelines, lender guidelines, reach out to my underwriter, get all the things done, but over here right now I don't really have to go ahead and wait for anything. I can just go ahead and ask that query over here to Mira and Mira can just go ahead and dig in all those guidelines and provide me all the information as possible. So as we can see over here that I have provided Mira with a scenario that during this credit analysis, this is what happened, and there are multiple derogatory accounts including a foreclosure.
(06:46):
Now Mira is pulling out all the information, giving you a step-by-step guidance, including the documentation requirement with the reference link of Fannie, me and Freddie Mac. So that way it has become so much easier for me that hours of wasting of time doing that research. I've just cut it short to just matter of seconds. Now that is the whole last but not the least. Today we have, this is one of my favorite tools actually, that's pre and writing. Just imagine when you're starting a borrower journey, tell a loan application stage. And today we all know how many underwriting touches we have on our files, which is an average of three to four or even five at times has gone high. Just imagine you're able to underwrite a loan at the initial stages and know what the profile looks like. So I have a borrower who's already completed his details, provided me with all the documentation required and then also gone ahead and I've already completed the product and pricing and I have completed the fees worksheet.
(07:47):
I have done all of those things and now I've gone ahead and pulled the credit as well. So with that, I can just go ahead and click on pre-writing and Mira will do a full finding review for me, helping me understand the complete personal details of the borrower, the name consistency through all the documents, the address consistency, the income analysis for consistency, the asset verification. As we can see over here, it has already identified there is an asset. So it has already provided recommendation of large deposit in the asset. So we just have to get that reviewed. It has already verified all the three Cs that is required and provided me a full very good underwriting findings. That has given me a very good understanding of what kind of a loan is this and how fast I can close this. Now just imagine this kind of a loan going into underwriting today. So just imagine the number of touches it can come down to. So with that being said, I'm so sorry, but we are not able to show you enough today, but there is so many more features of Mira that we would love to demo it to you and we would love to connect with you and show you along with that. So if you don't mind, you can definitely scan our QR code and feel free to connect with us later and we'll be happy to display it to you.
Bhagavath Bodagala (09:09):
And it's developed on latest tech stack so you can integrate with any of your existing applications too. Yes, thank you.
Christina Randolph (09:21):
So I know you had a bunch of hiccups. I think you overcome and recovered really well. I think this for me is really good example of AI starting to be infused in that upfront lending process. I think you hit all the right problem statements. It's the tragedy of the lending experience I think that we all know and appreciate today. Very clean interface. I think one of the things I like is how the guidelines are embedded. I would've liked to hear more about I guess the trust factor with the answers and the information that it's giving back. I think as we all know, that's going to be a hurdle with this type of technology with underwriters trying to trust something without having to do the research again. So I think well done just probably needs a little bit more exploration.
Bhagavath Bodagala (10:17):
It's
Christina Randolph (10:18):
An exciting time for this kind of technology.
Bhagavath Bodagala (10:19):
Yeah, whenever Mira answers, it gives you reference links where it is pulling information from guidelines. Yes,
Robin Clayton (10:25):
I was a little confused if this was a POS or if this was a loan officer helper or the guidelines there. I'm assuming the AI is closed loop if we're doing that kind of stuff. The other thing is that when you're showing the consumer some of the guidelines from Fannie Mae or some of that language, I would think most folks wouldn't understand that I don't know what DTI is, I don't know what Fanny is, what's a guide? Why does this matter? So just being very mindful of the persona in that this is a massive problem to solve. I'm excited you guys are taking this on because the potential is huge.
Panel Member 2 (11:02):
Thank you. One thing I appreciated was that the income analysis, if I interpreted this right, just one example is integrated and so if somebody is making this choice then it's not extra vendors. That stuff is also kind of built into the mirror program overall yet. Okay, so that part I liked. I do understand that you're trying to do a lot with showing the consumer and the LO persona and both with AI and having it get hung up and that's fine, but I had a comment about having the AI speak to the consumer in the correct way and I also had a comment about making sure that the inquiries, I was curious if the inquiries were pegged to lender investor guidelines and not just GSE guidelines as well, if it's in a particular shop. Is that a yes in terms of and it's just nod your head, it's not an open, it's not an open question, it's just nod your head yes or no if it's guides, non-agency guides in addition to agency guides. Yes. Okay, cool. Thank you. I know we're a little over too. Thank you. Thank you. Good job, Mira. Thank you. Thank you very much.
Panel Member 1 (12:04):
Great job Mira Labs. Thank you. Okay, next up Patrick Quast. Chris Fitzpatrick at Finastra.
Patrick Quast (12:16):
Thank you. In a world where the pace of life accelerates by the second, speed is certainly an asset that it's the harmonious blend of swiftness and precision and steadfast dependability that sets the truly exceptional apart from the merely fast. Hi, I'm Patrick Quast and with me is Chris Fitzpatrick. Patrick. We are with Finastra's mortgage bot LOS, and we are thrilled to be unveiling our cutting edge workflow automation solution designed to enhance your efficiency, precision, and consistency. Today's highlights will include our new UI and Lone Pipeline screen featuring customizable views that streamline and simplify tasks. We'll also explore our lights out integration with Lone Logics, utilizing advanced machine learning and rules-based systems to ensure swift and accurate data extraction and verification. Our commitment to innovation has led to the refinement of our document processing technology, prioritizing dependability when it counts. Chris will now showcase how these advancements will transform the mortgage experience.
Chris Fitzpatrick (13:24):
Thank you very much Patrick, and let's get right into the demo. Welcome to Mortgage Bott, LOS. We're showing. The first thing we're showing is our new UI/UX, which is a responsive and fluid user interface using a simple browser hosted by Finastra on the Azure Cloud. Mortgage bot can be accessed anywhere, tablets, laptops, or larger devices. What we're looking at here is our new loan pipeline, which allows users to create customized loan pipeline views that improve workflow and the ability to search access and configure how they want to do business. By creating simple pipeline views, it makes it easy for me to find my loans as well as to configure columns and filters that help me design the way I want to see my loan pipeline.
(14:22):
A little nervous up here. There we go, going into a loan. This is our loan dashboard view. Very easy to learn and use with simple navigation. Here on the left hand side, everything from all our processing and task views are easy to use. 10 0 3 screens, access to third party services including document packages through our partnership with DocMagic as well as all third party service integrations accessed, oh, sorry, I'm so nervous accessed within the browser. There we go. From ordering credit appraisals, MI title, automated underwriting, mortgage insurance and other services. Our intent in presenting this user interface to users is to keep all users within a browser so they don't have to leave the browser experience to chase down extra third party services or other functions within the loan processing experience. The second thing we're showing is our integration with our partner loan logics. We have a lights out integration that utilizes our built-in document content management system through our partnership with AxeCorp.
(15:46):
So anytime a document is uploaded through the borrower experience in our online POS or through APIs, they are immediately submitted for document processing with our partner loan logics. Now within document processing, what we do is automatically extract the documents from Loan Logics and submit them from image flow and submit them to loan logics for document processing. That includes document classification and data extraction as well as population of the data that's been extracted into the LOS. When I open up one of these documents to process, what I like about this is that the classification is already completed. The document is displayed side by side. It allows me to view information that was extracted and in this case I've already processed this document and shows the data that's been pushed from the document that was extracted and pre-populated into the LOS. This eliminates data entry improves accuracy as well as speed.
(16:56):
Clicking onto the next document in the workflow, this is a document I haven't processed yet and you can see more of what's going on here. So at the top here you can see the DA document has been classified associated with the borrower recognized as a checking account and associated with this asset within the 10 03 data entry. Now this could be this LOS data could have been entered by a loan officer or simply imported from the online POS and the data extracted is compared side by side, so it speeds up that stare and compare. It allows me to view the form itself, view the data that's been extracted, and with a click of a button it will populate and push all of the data that's been extracted to update the 10 03 without having to manually enter any of that data. Clicking next this case, this is an asset that is not recognized, so I simply associate with the borrower and identify that it's a checking account and here I can see that this Chase account is actually not been entered into the 10 03 before, so I really have nothing to compare it to. In this case it is a new account, so I can simply create a new asset from the document itself. So it extracts the data from the form pre-populates the 10 0 3. So I simply view the document, compare the data, save next and eliminates all that data entry and approves accuracy and speed.
(18:32):
The third thing we want to show you today is rules automation using the appraisal that was again uploaded into our document management system. It's been submitted for rules automation. You can see here there are just a few rules that still need to be evaluated. In this case, these rules have not been able to be evaluated because we still don't have the contract of sale to compare to. If I flip this switch here, it also shows me all the rules that have been evaluated, all the rules that have been passed. Again, the rules that have been pending that I need to work on as well as rules that do not apply. So these rules again through our partnership with loan logic are really kind of pre underwriting rules brought into processing to help streamline and improve accuracy and to reduce buybacks at the end of when the loan is closed and sold. Also, lastly, everything is tracked within an audit log. So going back to look at one of these loans I've reviewed, everything that I've tracked that I've done is tracked for the end user for transparency and workflow automation. With that, I'm going to turn it back over to Patrick to wrap up our presentation.
Patrick Quast (19:50):
Thanks Chris. The mortgage process is like a race with all lenders looking to close loans quicker. Our workflow automation solution will pull you within feet of the finish line. Again, I am Patrick. This has been Chris, we're with PRAs mortgage bot LOS, and if you're not just looking to be fast but be exceptional, please stop by our booth 15 to learn more.
Chris Fitzpatrick (20:10):
Thank you very much.
Christina Randolph (20:17):
That was a very dramatic opening. Patrick. Good job. I will be honest, I've seen lots of OSs. I didn't see anything super groundbreaking here, but I think the way that you have redesigned and reformatted your UI because I have seen your solution in the past, I think is really critical for adoption and pull through and to be able to achieve what I think you're trying to achieve with the workflow and operational efficiency. I think again, you're putting together lots of things that historically have not been able to have been put together and stitched together in a great organized way. So I think for that it was a pretty good job.
Patrick Quast (21:02):
Thank you. Thank you.
Robin Clayton (21:03):
Patrick, can you step a little bit to the side?
Patrick Quast (21:07):
Oh, like here,
Robin Clayton (21:08):
Look at these outfits guys.
Patrick Quast (21:10):
Yeah,
Robin Clayton (21:10):
Amazing, amazing. If I was scoring on that, 10 out of 10, really interesting. I love the slide out design. You truly spent a lot of time on this UI and UX, so really, really good job. If that was a mission home run for the rules, it was showing pending, pending, pending and then verbally you said it was missing the contract would've loved if it had told me that's why it wasn't passing. And for the side by side the steer and compare, I still had to read it right? Did it match? Did it not match? So it would've been cool if it maybe had some color coding or maybe it only highlighted the things where there was a discrepancy. It would've been cool to see that, but yeah, absolutely incredible. I love you. Shout out the partnerships API integration, always a fan and you both look fabulous. Thank
Chris Fitzpatrick (21:57):
You so much.
Panel Member 2 (21:59):
In terms of go to market questions, I think that integrating into your presentation, which channels that you're targeting and also a little bit of emphasis on how many loan types and what types of loan types the system can handle, I think that's critical, especially since your firm as a whole covers so many verticals. I think in the mortgage vertical it's like what are the channels, how many loan types? That's critical one. And then I did want to see more on status of the workflows. You talked about the workflows. I did like the UX too. The customization of the pipeline view is great, but workflow is what the LOS is all about and that's what I wanted to see more of. But good presentation, good UX. Thank you.
Chris Fitzpatrick (22:39):
Thank you.
Panel Member 1 (22:41):
Great job Finastra. Up next we have Todd Norton and Tony Farnsworth with Maddock Insurance Services.
Tony Farnsworth (22:54):
Good afternoon everyone. Appreciate y'all being here. We're super excited to talk to you today about how MAD is transforming the way people shop for homeowners insurance during the mortgage process. Why is this imperative? Let's talk about that a little bit. If you think about a first time home buyer coming in, getting a mortgage, the loan officer says, Hey, you need to have homeowner's insurance in order to close this loan. So what do most consumers do? They maybe go to a local agent or shop around there, go online, they get asked 30 to 40 minutes worth of questions. The agent has to come back and give them the quotes from a limited pool maybe that they have. They choose one and then it's still up to your team to chase down the documentation in order to clear that condition. And what we've done at MADD is we have streamlined that entire process within just a few clicks.
(23:49):
A consumer can come in, they can shop from our network of over 50 carriers real time. They can choose that quote, they can bind that policy and we will automatically deliver the documentation necessary back to your operations team to clear that condition. So very exciting to be able to do this. So let's jump into the demo and show you what this looks like. So the first thing that's important to note is we try to create a personalized marketplace for every consumer on every file. And so when we work with our partners, we share data in a compliant fashion, we bring that in, we create that unique marketplace for that consumer, and then we will take an omni-channel approach, whether it be via email, could be point of sale, could be through the processor to be able to give them the opportunity to shop for insurance.
(24:38):
And this example here, this is an email that was sent to the consumer. You can see that between MAD and a BC company, which would be the lender, everything is co-branded. On the page itself, we have a get homeowner's insurance link. The consumer would click that and ultimately go to the next page. Obviously with a lot of regulation changes coming up, especially with TCPA coming in 2025, getting the proper consents is really important for insurance. You have to have TCPA consent. You also have to have FCRA consent. We get that directly from the consumer. So once they click get quotes, it's going to take them to an Expedia style shopping experience for homeowners insurance. You can see based on here, the profile on the left is the data that ultimately was pulled in not only from what was shared from the partner, but also from third party data sources.
(25:30):
And then we have a list of carriers that based on the information that we sent via API we're willing to deliver a quote. So you can see safe code there has a $680 quote right below that nationwide at 7 54. The other cool part about this is you can click in to show details so the consumer can see dwelling coverages, personal property liability coverages, and they can even compare the two of them. So Todd, why don't we pick a quote. We're going to go with AAA as we move forward here. One of the next things that we're going to do is we're going to look for additional discounts. So if they have protected devices, if they have smoke detectors, fire alarms, what have you, those can add additional discounts to the premium. So we go through and complete the discounts. The next four buttons or so, we'll just confirm the information that we already have on the property.
(26:21):
Since we work with over 50 carriers, all carriers have different requirements for what they need in order to bind the coverage. We put that all together in one. We surveyed the consumers and we asked them, would you rather do that on the phone with an agent or would you rather self-serve online? Overwhelmingly the consumer said they would rather do it online. So as Todd's kind of going through and just confirming these are basic things, what type of siding, what type of roof do you have? Do you have a finished basement? Do you live near a fire station? Some of those types of things. So he's going to go through that once he gets to the end of all of these questions in here and we're just going to fly through it due to time. What our engine will do, which we built proprietary here at Maddock, we have a proprietary rating engine, which is somewhat similar to A PPE.
(27:11):
We also have a proprietary agency management platform for our agents to be able to use and deliver a great customer experience. And so what it's doing right now is it's going through and re all of those because based on how they answered those questions, there might actually be a better policy than the one they originally chose. So as you can see here, it came back and said, Hey, Safeco is actually better coverage than what you originally selected. You can go into the details and see everything that's associated with it. Let's say you really like AAA because you have your auto insurance through AAA and you want to stick with that coverage. You can ultimately do that too. Some carriers allow for us to be able to have the consumer go through the entire flow and buying the coverage online without ever talking to an agent.
(27:58):
We're one of the largest agencies and have the biggest number of carriers that allow us to do that. Or they have to schedule a call to finalize the policy with an agent. So we've made it super simple Calendly type of experience. They can go in and pick a date, they can pick a time. Once they submit that, that'll go into our agency management platform and one of our agents will call them at that specific time to help them finalize everything. Once that is done and the consumer has their policy, Todd will show here. This is an example of an email that we sent back to not only the consumer but to the loan officer and their operations team with all the documentation that they need, replacement cost estimator deck page, and then of course the policy invoice. So everything is done and taken care of. We can do that via email or other integration opportunities. So at the end of the day, they went through this entire process, they can do it in roughly five minutes, have their coverage taken care of, and that condition is now clear. So that's all I had for today. Thank you for your time. I appreciate it.
Christina Randolph (29:10):
Well, I love anything that helps further automate, but it's still sort of humanizing parts of the mortgage process. I think as a consumer, again, there's not a lot of positive feedback about someone's mortgage process. So wherever you can help create a better experience, there's always a positive. I would've liked to hear more. I think you had about almost a minute and a half left more about integrations into point of sales OSS pulling. What more of the lender would experience, can they get that policy? I think you mentioned it but would've loved to see how that maybe helps clear condition, maybe quantify how much time savings. I think we all know how much time a process like this takes, but it would've been nice to kind of see it together side by side of what you're solving for and what kind of operational efficiency a lender can gain. But again, anything to help create a better experience. I'm a big fan of, so good job.
Robin Clayton (30:10):
Yeah, insurance is a huge issue. That's been a hot topic at this conference for everybody, especially those of us listening in California. I love that you guys are tackling this space. To her point about the L-O-S-P-O-S integration, watching you do the demo and filling out number of beds and baths, the age of the roof would love to see. Maybe you extract that from the appraisal or other fields that are already have been input into the system. Why am I having to do that a second time? UX is stunning. The Calendly integration for scheduling the omnichannel people want to relay by email or text or in a system. So very well done.
Panel Member 2 (30:48):
I had similar remarks on questions. How does the EOI get into the LOS and get cleared? If the customer can't answer those technical questions, can it just be pulled from the appraisal? So I didn't see that in the consumer experience, but I just want to underline the Robin's point 50 insurers right out of the gate in this truly consumer grade UX in a market like we have now where insurance is by far the hottest topic for a consumer, they have the realtor to manage, they have the loan officer to manage and now they're like, oh God, I got to get insurance. I don't even know if I can depending on where I am. And this to me is one of our categories is market potential, market relevance. And right now this meets the market and I'm impressed by that. Good job Mattock. Thank you guys. Thank you.
Panel Member 1 (31:40):
Okay, up next we have Marc Hernandez and Han Oksuz with Guideline Buddy.
Marc Hernandez (31:47):
Thank you. Good afternoon everybody. Before we get into the juicy stuff, let's do some fast facts. So the average American adult reads at about 200 words per minute. So if you have TOL soy fans out there, if you've ever read War and Peace, it's going to take you about 49 hours and that's because it's got almost 600,000 words in it for all you Game of Thrones fans out there, the George a r Martin series, the Song of Fire and Ice has over 1.7 million words. That's 144 hours of reading, but that's a series of novels. And why are we bringing this up? Because guys in our industry we have a series of novels. They're called underwriting guidelines. They are Fannie Mae, Freddie Mac and all the HUD handbooks. And if you combine them all, it's over 1.8 million words. That's 150 hours of reading. That's not including all the bulletins, that's not including all of the mortgage e letters.
(32:42):
That's an overwhelming amount of data and text. And we don't get paid in the mortgage business to read Game of Thrones. We get paid to read these guidelines and if we don't search them properly, if we don't apply them, if we don't interpret them correctly, we make costly mistakes. We've all done the control f search that PDF to find what we're looking for. And if we don't do it right, that mistake can cost a lot of money. So guys, what do we do? What is the solution? Well, today we'd like to introduce you to Guideline Buddy. We are a generative AI chatbot solution that helps mortgage professionals not only search but interpret mortgage underwriting guidelines with our AI model specifically designed for the mortgage industry. So now the juicy stuff Han, let's go ahead and queue it up here. This is guideline buddy and we're going to start first with a very simple question that we all know the answer to, but it'll give you a sense of what it is. Han over to you.
Han Oksuz (33:48):
I already set it up. The first question we have is, can a borrower make a down payment with money that is gifted to them?
Marc Hernandez (33:54):
So what that question went into is we chose the 4000.1 FHA handbook and the simple answer to that question is, yes, of course you can use gifted funds, but imagine the mortgage professional, maybe a loan officer asking their very busy manager, can I use a gift to qualify my borrower? And it's like, oh yeah, go ahead and do it. But now look at guideline buddy's response. This is a very comprehensive thorough answer. Not only is the answer yes, but it is covering everything in the 4000.1 that is relevant to gifts so much. So if you scroll all the way down, you can see that there is a source link to page 1104 of the 4000.1 that begins to talk about gift funds. Easy question. Let's move on to something a little bit more complex perhaps a question for Freddie Mac and perhaps a question that may take a veteran to actually download the selling guide or find a way to search it. And I'm going to kick it back over to Han to tell you the question,
Han Oksuz (34:58):
What does a borrower need to qualify for a delayed financing exception when refinancing a property bought with cash?
Marc Hernandez (35:07):
So a lot of refinancing is going on here as rates start to drop a lot of equity in the homes. So again, guideline buddy went to this Freddie Mac sellers or servicer guide and we have a technology called Ag agentic ai. You may have heard that buzzword coming out here. We have a series of agents going in to the Selling to Servicer guide, pulling all relevant information for delayed exception financing and giving you a very comprehensive answer if you're a production assistant, a loan officer, processor, underwriter, a post closer, a scenario desk. If you're touching guidelines guideline, buddy is made for you as a super assistant, I'm going to geek out on some Game of throne stuff. If you've seen it, you have the hand of the king or the queen giving you that advice that you need to make the accurate answer. Now let's take it up another notch. Instead of asking it a question, why don't we give it a scenario? Did you know that there are Facebook groups out there with tens of thousands of mortgage professionals that put their scenarios out there for the world or the mortgage world to answer for them? That's crazy. It's people relying on people, but what kind of authority do you get? So I'm going to have Han read you this scenario.
Han Oksuz (36:20):
Client purchased and lived in a home while on duty, then hired a property management company to lease out in May of this year. They're currently living with family and while waiting to purchase the new home where they will be moving after separation underwriters saying we can't use the rental income because we don't have two years history, I thought we could use it to offset the mortgage payment at least.
Marc Hernandez (36:46):
So this was a real question and the person got their answer, but how long did it take? Right? So now you have a very comprehensive answer here, and guess what? We made the loan that the underwriter didn't want to make and now we have a happy borrower, we have a happy loan officer, we have a happy company. It's a win overall. So our strategy at Guideline Buddy with our tool is to give access to mortgage professionals to this technology to boost their productivity and to lower the cost of origination and make more loans possible. We're also on a mission to democratize access to gen ai. You've seen a lot of gen AI solutions today, but we have to adopt it as an industry to make it powerful. We're so committed to this that we have a big announcement to make and we're going to ask guideline buddy to reaffirm that decision. Han
Han Oksuz (37:39):
Hey buddy, the leadership team at Guideline Buddy is considering making its free for all mortgage professionals. This would help them work faster and more accurately while learning to use generative AI chatbots. Do you think that's a good idea?
Marc Hernandez (37:56):
So if you're reading the screen, I'll give you the excerpt here. Guideline buddy. And by the way, this is not planned. This is generative AI genuinely giving us an answer guideline. Buddy says it would. Making it free for all mortgage professionals sounds like a great idea. Give me the screen right back, okay. It will definitely help them work faster and more accurately. So starting next week, every single mortgage professional can go to guideline buddy.com and subscribe and have access for free to search all agency and HUD guidelines. Now we are very dedicated to pushing adoption of this technology. So we have enterprise clients, we're going to be outside in the booth there if you want your very own guideline, buddy White labeled. We build chatbots for mortgage companies to use for internal purposes, but we want to push the industry to use this technology and democratize access. We got one minute left and I am again, if you can't tell a big Game of Thrones fan. So Han, let's go ahead and give it one more question.
Han Oksuz (38:58):
Well, a guideline buddy asked me what do I think? And I said I think it's a great idea, but to clarify it a little bit further, I said, explain it to me if we were in the Game of Thrones.
Marc Hernandez (39:08):
It says, ah, as a fellow mortgage professionals seeking wisdom in the realm of guidelines, imagine if you will, that navigating mortgage guidelines is akin to traversing the treacherous lands of Westeros in the vast kingdom. You are the loan officer, you are the noble knight tasked with duty of securing the best outcomes for your clients. And it goes on. We just wanted to have a little fun with GenAI. It's a very, very powerful tool. We want to put it in the hands of mortgage professionals to make everyone faster, better at what they do. And we want to thank you for your time. We want to thank the coaches and the judges. We appreciate it.
Christina Randolph (39:48):
Well, I'm not a Game of Thrones fan unfortunately, but I still got what you were saying. I think like everybody has been saying Gen AI solutions, while I think everyone's excited for the emergence of it and the infusion of it into mortgage, I would've liked to probably hear more about the risks and the guardrails around the answers that it's giving back. Maybe a little bit about how this would be different than a Google search, which pretty much everybody does today. And then I would've loved to have heard a little bit more about your integration distribution strategy. I think it's a great tool, right? There's no doubt that there's going to be multiple thousands of use cases for this kind of functionality, but just how you're going to go to market would've loved to hear that a little bit more, but well done.
Marc Hernandez (40:36):
Thank you.
Robin Clayton (40:37):
Their go-to market is everyone gets it for free. That's pretty good. Really interesting. This literally is a problem we're having right now, a sales manager, Hey, we're looking for guidelines on Facebook. There's got to be an AI solution. So yes, you are absolutely on the pulse of what LOs are thinking about. Would've loved to see an omnichannel solution to this getting LOs to download an app, go change their process. They are addicted to email. So if there's a way to do this via email also, I'm assuming when you mentioned the enterprise solution, you could put in lender overlays and you can customize and have a little bit more protection around this and just a nitpicky thing, making it full screen, maybe larger font for those of us that our eyes are going bad and the ability to copy. So if I wanted to copy and paste and send it to an underwriter and say, nanny, nanny, boo boo, look I showed you. They said it's true. So really cool. Love the generous spirit of involving the community. That's incredible.
Panel Member 2 (41:36):
Thank you. I'll pick up there. It seems to me that's also a clever way to train the model. So I do quite like that idea. I will give you a 10 out of 10 if you can have it, call your every loan officer your grace. But no, I actually really do like the natural language. I like how it behaves. You can ask it questions as a loan officer, like I'm talking to my team in the office. Or there are forms on social where there are guidelines. I presume if it goes free for the space, that kind of then ends up replacing it pretty fast. So the question that I had, and I think you alluded to it with the enterprise uses, so just nod your head if this is correct, with enterprise uses, you can put in your own guidelines as a lender instead of just the Fannie Freddie, FHA, VA USDA, correct? Yes, absolutely. Okay. Alright, thanks. Good job. Guideline buddy. Good stuff. Really, really cool. Thank you everyone. Thank you. Great job.
Panel Member 1 (42:40):
And next we have Richard Wigton with Optimal Blue.
Richard Wigton (42:47):
How's it going everyone? Yeah, my name is Richard Wigton. I am a Senior Hedge Account Manager at Optimal Blue. I'm hoping that you've heard of us a little bit here over the past couple of days, but please feel free to stop by the booth after if you have any questions on the presentation or any of the other products that you see here and that you've hopefully heard of today. We're going to be focusing mainly on the hedging and trading aspect of our services and more importantly, our new hedging software platform, compass Edge. So this product will be for anyone who is currently hedging their loan pipeline or plans to in the future. With Compass Edge, we are aiming to maximize lender's profitability with risk management tools as well as dynamic functionality. We are aiming to reduce manual processes that we've grown accustomed to over the past years, giving people their time back and really just streamlining the entire thing here.
(43:43):
Most of these manual processes are associated with profitability, research, projections, trading, just day-to-day activities and things like that. And then we're hoping to tackle all of those could really speak for hours on all the exciting things that we've released in the past little while here as well as what we have upcoming. But today we're going to focus on just a couple things. First off our integrated dashboard that we have into our site as well as a couple of the AI tools that we have recently released into the platform. So without further ado, there we go. So yeah, this is going to be our dashboard screen that I mentioned. Goal of this page here is really for anybody in the company to be able to jump in and have exactly what they need at a high level at the forefront, whether you're a loan officer, C-E-O-C-F-O, secondary market manager, really just bringing you everything here on this front page from day-to-day numbers like net position pipeline totals, and then getting deeper down into some shock profiles around gain and loss numbers as well as some investor color all built in and integrated within the site, clickable and everything like that.
(44:55):
Hoping to really just make it easy for everyone to use quickly, whether you're about to run into a meeting or sitting down for breakfast, just having a quick look and getting a quick picture at where things are on today. And then to continue those themes, right, easy to use, hopefully saving you some time. We introduced some of our generative AI tools. We are going to be focusing on three of those. The first one is going to be our profitability assistant. Second one's going to be our projections assistant, and then we have our trade assistant as well. The profitability assistant is probably my favorite personal favorite here strictly because I think this is one of the hardest things that secondary managers and lenders have to do on a daily basis, which is explaining daily movement of gain or loss in your pipeline. Traditionally it's something that takes an experienced capital markets professional to comb through numbers of reports and data sets and then use intuition to come up with a narrative that makes sense.
(46:02):
Obviously can take a couple minutes on some days, but in a volatile market or on an off day we could have a couple hours digging into reports and other things like that. And that certainly can add up. So what our tool is doing here is it's going to take the customer's hard data and it's going to ingest it and then create a digestible executive style summary about what may be going on in your pipeline on a day-to-day basis. So again, we have the summary here. We also have, we give the top five drivers to what may move a hedge pipeline on a daily basis, helping you to kind of drill down maybe some patterns that you've seen or just getting rid of any one-offs.
(46:49):
The next one that we will go look at is going to be our projections assistant. And what we're doing here is same thing, taking your data, generating an AI statement, but what we're doing is estimating or projecting certain things, and this is going to be something that's going to help answer questions that are asked to themselves as lenders or secondary managers on a daily basis. What will locks look like today? What will my loan sale look like today? Anything like that. We've taken that data in and we've blown it out to show a predictive model based off of any time of the day, anytime of the week. And these numbers, they're not just numbers, they're also built into the site to where they're able to be applied. So we can put certain things into this model and immediately review that impact back in our position model here.
(47:46):
So now that we have our new position, our projected position, the next step would be to go and use the trade assistant. The trade assistant is using hundreds of data points, looking at market liquidity down at the coupon level and just suggesting the cheapest and most efficient trades here for the secondary manager hoping to aid and assist them, obviously not replace them just here to help the aid and the clause towards making sure that your position is hedged as efficiently as possible. From here we have the trades at the bottom. If you do like what you see and you like the changes, you can quickly add them to the blotter and that will go down to the bottom. And we'll also adjust your hedge position. So now with a couple of button clicks here, we have completely readjusted everything with use of AI and kind of takes that daunting task and then turns it into something a lot easier and just really making everything quicker, cheaper, more efficient, and hoping to give you some time back there. These capabilities have multiple benefits, efficiencies and things like that. Saving people time, but also the ability to save companies basis points and that's obviously important here. Yeah, that's all I had today. Thank you very much for your time and please keep an eye on us. Thanks.
Christina Randolph (49:26):
Well, I think I understand. I think you might've in a last minute addition to this demo, so I think you did a great job for kind of getting thrown in last minute. I think any tools that help lenders track, monitor and find opportunities for profitability, they're huge in this market. I like the look and the feel like I try to use dark mode wherever I can just to help with the eyes. So it's a nice Bloomberg s terminal fields, which is nice. I would've liked to see a little bit more. I know you got thrown in last minute, but I think great job, good luck. I think it looks like it's a good user interface, easy enough for even a business user to use. Thank you. So great job.
Robin Clayton (50:14):
Yeah, I agree with Christina. The visualization was great. We'd love to see more folks go to dark mode for sure. We've been staring at screens all day. One of the things that was really exciting about the profitability assessment is the visualization. Being able to see that and that sort of bar chart with the red and the green. And whenever we're dealing with quantitative data and people who are moving really fast and they're tired, they're getting distracted, that's when you see a lot of human errors. The idea of AI saying, Hey, I'm going to type in 2.214 million versus someone trying to fat finger that or type errors. I think there's a great potential there to be a lot more accurate with that. I am curious though of a world of, if a lot of people use the trading part of it, if it would actually start becoming a beast on its own and sort of eating itself. Sometimes you see AI models get a little loopy, but really interesting and I love your closing statement. It's about making money. It's about profitability. That's what we love to see. Job well done. Thank you.
Panel Member 2 (51:15):
I said this yesterday in the opening keynote session, but gain on sale is the lifeblood of our whole business. So that's one. And I think therefore from again, market relevance standpoint, it couldn't be more relevant. And then two, I think that being able to provide technology, not just to sales and fulfillment, but when when we think about your category, we often only think about PPE. The thing I appreciate about OB is that you go from the first touch with the consumer, obviously through sales and fulfillment and all the way through to capital markets. And this is a look that people don't always get of what are the capital markets desks using to see? That is something that I like to see in how gain on sale is being managed. I didn't understand if it's gen AI, but I would imagine that gen AI couldn't be being used on numbers that are so precise if gen AI is by definition of black box. So I would've appreciated a better explanation on that, but I understand that it is doing some analytics to help the trade assistant as an example, but I just wanted more on that. But everything else I said holds great job Optimal Blue, everybody.
Demos: MIRA Labs | Finastra | Matic Insurance Services | Guideline Buddy | Optimal Blue
September 27, 2024 11:54 AM
52:38 Sponsored by
