As Blooma’s Director of Customer Experience, I work a lot with lenders and brokers, especially their commercial real estate underwriting teams. I’d like to think I’m familiar with the motivations and struggles of our clients, since in a former life I used to be a CRE underwriter as well. The truth is, CRE underwriting and analysis processes are very fragmented, which leads to inefficiency.
Much of my job as an underwriter involved the management of many disparate data sources. For as large and mature as the CRE lending and investment space is, underwriting processes and innovation have not evolved or kept pace.
Got a new deal coming in? You’ll start by pulling up a model you used on a prior like-kind deal, and then overwriting the data. You then begin searching for sales and lease comparables on multiple sites, speaking with brokers about transactions and trends in the area, reviewing maps and street views to analyze the area surrounding the property, rounding up and reviewing title data, creating proforma cashflow models, the list goes on. Not to mention requesting and analyzing borrower financials, all in different mediums. You become a data aggregator, pulling in files, documents, and data from numerous sources and saving them all into a shared folder, drive, or system where they come together and start to make some sense. All of this work ultimately culminates in a final Excel model and a credit report/memo drafted in Microsoft Word. This is a painful process, and no matter how good you are at it, by its nature it is prone to errors.
An underwriter’s true job is to come to a binary decision as accurately as possible. Are we funding or passing on this? That’s it. But don’t get me wrong. Underwriters do not have a simple job even if their end goal is to make a decision one way or the other. And the “noise” of the process described above doesn’t help to get there – in fact I think it can mostly distract from this goal.
I think technologically we’re at a bit of a crossroads. Advancements in artificial intelligence have made it possible to take the more monotonous and time-consuming parts of the job of a CRE underwriter and hand a lot of this to computers. Tasks like parsing through borrower, financial spreading, searching for comps, etc. This allows underwriters to be much more efficient, and frankly frees them up to review each deal more analytically. To use their vast experience to engage in a strategic conversation about a deal and whether it should happen, which is what they do best.
If we hadn’t developed this approach at Blooma, someone else would have because its time has come. The only question now is how quickly lenders and brokers will embrace it. I’m excited working with all of the lenders and brokers that already have. It feels good to make the jobs of CRE professionals a little more enjoyable, by helping transition them to a newer way of doing things. It wasn’t that long ago that I was one of them.
I demonstrate our product to many people, so I get to hear a lot of feedback and questions. One of the most potentially confusing aspects of what we do is using artificial intelligence and machine learning to generate a “score” for a certain commercial real estate property or proposed deal. It’s sometimes hard to explain what this really means, so I’ll try again here.
People sometimes think of the scores that are awarded an Olympic gymnast after a routine – we’ve all wondered about what can sometimes appear to be an arbitrary system of judging something. But we have to remember that this kind of interpretive scoring of a performance is not what is happening in the world of artificial intelligence.
When we engage with a CRE lender, our AI technology learns their own specific ideal lending profile. Over 3,000 data points are summarized into a “score” for them. The real value is in the automated collection and analysis, of course, and the score itself that is generated at the end is really just an interpretation. A computer can keep track of 3,000 data points and see the result in the data, but a person needs it to be streamlined and summarized. The artificial intelligence doesn’t start out with any idea of what is a good or bad deal, it simply compares all of the data to the lending profile that was preset by the user. By learning what kinds of deals you like, in other words, it can tell you how good a new deal should look to you. Our different customers might score the same transaction in wildly different ways, and perhaps they should.
Often times, our customers want the score explained to them. This makes sense – if your child came home with a C on their report card, you might ask them why it is a C rather than a B (or a D). The score itself doesn’t tell the whole story. This is one reason we use the type of machine learning that we do, so that we can go back and explain every data point that led to a conclusion. In other words, we can tell you exactly why your child got a C instead of an A. Some deep learning models (neural networks) don’t really allow for that kind of going back and parsing out exactly where each piece of data contributing to the whole originated. This would obviously not work in regulated environments such as banks where an audit trail is always necessary.
Finally, there’s another advantage to teaching AI what you like and then letting it score opportunities for you. You may think you know all of the factors that determine whether something fits your ideal profile, but sometimes the AI gets to know them even better than you do. Data doesn’t lie, and I have definitely seen examples of an AI system being able to “teach” users that there may be certain patterns that they are looking for (or avoiding) that they weren’t even aware of. For example, AI might look at a large number of previous loans and be able to find patterns in the data that led to late payments or even loan default, that you might have not been aware of. Now you can simply use this insight to update your lending profile. It can be disconcerting to learn that the way you’ve been articulating your ideal profile might not be as accurate as you thought. But those of us who are open to learning something new can benefit from this greatly.
In a collaboration between BCG and MIT from 2018 (Ransbotham et al., 2018), researchers found that when it comes to artificial intelligence, companies and organizations can be classified into 4 groups: 18% Pioneers, 33% Innovators, 16% Experimenters and 34% Passives.
Pioneers are enterprises with an extensive understanding of AI tools and concepts, and embrace AI in significant ways. Innovators have a good understanding, but still display little actual application of AI in their business. Experimenters are using AI for their business, but without seeking an in-depth understanding of the AI methods. Finally, Passives lack both in-depth understanding and application of AI technology.
Interestingly, all four groups agree that AI will change their business model in the next few years. This means that sooner rather than later, AI applications will penetrate the entire corporate landscape. This will be important as AI becomes a pillar of competitiveness, allowing companies to get things done faster and more accurately and to reduce time spent on less desirable work.
In the past, laws around data privacy have hampered the evolution of AI applications in many industries which deal with private data, such as sensitive or personal information. Typical examples are the healthcare and financial sectors. In recent years however, emerging technologies have enabled the secure use of private data to build and use AI. These private AI technologies are accelerating the quantity and nature of novel AI applications, and transforming the AI landscape. Thus, Pioneers, Innovators, and Experimenters in the financial sector have started to use AI on their private data.
Private AI technology allows a company to use data to train highly performing models without exposing or sharing confidential data. Although the field of private AI is very cutting edge, there are a few notable methods already, including differential privacy, homomorphic encryption, federated learning, and data anonymization. For example, federated learning is an advanced, state-of-the-art method used by Google and the healthcare industry (full disclosure: Blooma uses it too). This method involves making use of client data, without the client having to share their data with anyone.
In a typical set up, a model is deployed to those client locations and trained on the data there. After training, the model is sent back to a central location where it is merged with the models of other clients. After this merge, you end up with a model which has learned from data across all clients without sharing the data in one central place. Federated learning, if done right, ensures that there is no exchange between each individual client’s data. Thus, federated learning allows a company to learn from confidential data while keeping the data secure and private.
Meanwhile, the introduction of AI into the corporate landscape has led to changes in legislation which continue to develop. The General Data Protection Regulation (GDPR) and California Consumer Privacy Act (CCPA) are two statutes that have recently come into play in the attempt to promote and regulate data privacy and security in the European Union and California, respectively. Additionally, Brazil passed the General Data Protection Law that came into effect in February 2020. According to Richard Koch, Managing Editor of the GDPR, EU, these legislations that have begun to emerge have some common principals, including the importance of defining personal data and certain fundamental rights relating to data subjects (2018). Private AI allows us to obey those laws while capitalizing on private data. Whether you consider yourself or your business closer to the Pioneer or the Passive end of things, your data is already a part of this changing landscape, and security will continue to be the driving factor in the future evolution of private AI technology. Despite the many differences between the statutes mentioned above, we can only assume this is just the beginning of the movement toward an increase in legislation to govern data privacy.
Selections from a Q&A focused on Strategic Planning In A Rapidly Evolving Digital Environment.
Al Dominick, CEO of Bank Director, recently put some questions to our CEO, Shayne Skaff.
Al: What have you noticed post-PPP about the banks’ interest in bringing in new technology ideas – what have you learned over the last few months having heard from your clients about what they went through but also what they’re aspiring to become?
Shayne: Obviously, origination stopped in March, so it was really about asking how is their current loan portfolio being impacted in the space. As a technology company, we’re helping them understand with their existing portfolio, what does this new market look like for them, how are these assets that they’re managing being impacted by this new environment, and then our system is analyzing assets every day and our banking customers can see how the environment is actually impacting the current loans that they’re holding on the books. So, as the people were pushed to handle the PPP loans, the technology kind of took over, and it was still showing exactly what status their current business was in during this difficult time.
Al: As we think about planning for the future and the need to be agile, how are you starting to see banks frame that concept to solve specific business issues?
Shayne: So much of agility comes from the ability for humans to make quick business decisions. Data isn’t static. It’s constantly changing, and it’s important that we as humans are seeing the results that are taken into account with this dynamic state of our business. A good decision today can turn into a bad one tomorrow unless you are actually managing to that dynamic data on a real time basis. What we’re helping our customers do is really present the results of what real time data means for their business, so that when they see those results, they can make decisions on the spot.
Al: When you think about individual expectations and how they’re shifting, help us to contextualize AI as it is now being used to address some of those customer confusions, or those things that are happening where you really want the machines to be crunching the data to get a better picture of the world in which you live.
Shayne: When most people think about AI in the workplace, they don’t initially think about the customer experience, they think about replacing jobs within companies, which is an easy place to go to. The way we think about AI is as a way to actually improve the employees’ work experience. We ask how do we move the employee to the art of their job and use the machine to do much of the science, really the work that the humans don’t like to do. If we can come in and use the machines to do a lot of the tedious work and number crunching and move the human to the relationship and the art of their work, that’s really where we’re focused. The way we look at AI is from the perspective of how to improve the workplace, and that generally translates into speed for the customer.
Al: When you’re helping banks think about what is possible, how much of a goal could you realistically put in front of them for the next 12 months or the next year and a half?
Shayne: When you have a heavily regulated industry like commercial banking, the first question is can I do that safely and will I be breaking any regulation. From a technologist’s perspective, you have to be patient in taking the bank through the process and help educate where we’re at with the technology, what it means from a security perspective and that everything is safe. All you need to do is look at other industries to see where banking needs to go because there are others that are so much further along. We’re going there, it’s just a matter of time and willingness to take those steps. Not coming out of traditional banking, I was told going in when we launched Blooma in 2018 that it was a very difficult space to sell technology into, but I’ve actually been pleasantly surprised with the reception that the banks have given us, not just community and regional, but some of the largest national banks in the country. I think there’s less fear now, and with every month there is less and less fear as people start to really understand that they need to produce speed, their customers are going to expect speed.
Al: Should early stage fintechs be judged by a different measuring stick as compared to established or traditional longer term vendors when it comes to vendor management?
Shayne: It’s super easy to implement a cloud solution now. It’s not like you have to pay a million dollars to implement and a huge on prem licensing fee. With a lot of providers you can try before you buy – they can launch their platform and allow the bank to get comfortable with it for 30 or maybe more days and then go into a commercial agreement. The flexibility with how technology is being consumed and purchased is like never before, it’s a whole different game today. As an early stage tech company, we’ve been super flexible with our clients in terms of how we engage with them commercially, the reality is that we’ve taken really all risk off the table. We know we don’t have a decade of customer references behind us that they can pull from, so just as a vendor make it super easy, take as much of the risk off of our clients hands as possible, allow them to do things like pilot and just not make it such a stressful decision for them.
Al: What helps them to say I’m doing something I’m comfortable with, but now I know I need to shift gears to something I’m not as familiar with but I think is the right thing. Is there a trigger that you’re seeing that others might appreciate?
Shayne: For us, we really try and come at a new client with how we’re going to use technology to improve their employees’ lives, which ultimately will resonate in better service for their customers. The workforce is changing, the people that have been doing the underwriting and all of the analysis for the last couple of decades are now moving out, and there’s a new workforce moving in that wants something different and so we generally talk about that first, that human element, and that seems to work well.
Al: How are you envisioning the first 6 months of 2021 shaping up?
Shayne: I’m generally an optimist, I won’t try to do this from a banking perspective but will do it from a technology perspective. I think the market itself has some big opportunities in my world on the CRE side and some extreme difficulties, and it will take more than a year to see where everybody lands. From a technology perspective, we’re seeing the end of monolithic technology. There’s so much modularity that exists out there with cloud platforms and the ability to plug in to various systems and integrate with various systems, and being able to utilize all of the data that exists out there to provide very quick decision making for business is going to be super important. Where AI and ML and some of these technologies are going to take us is really to a place where we’re going to have answers before we’ve even asked the question. There’s an answer for me in my computer right now to a question I have not asked yet, and that’s where were going and we need to embrace that because that’s here in 2021.
You are standing in line at your favorite deli and you see that they recently changed the entire menu. There is a long line of people behind you. You feel a slight increase in your stress level as you tell yourself that you have about 20-30 seconds to decide what to order, to be considerate to those waiting behind you.
Your analytical mind immediately strikes out a few items for being too expensive or not in line with your diet, but you are still left with quite a few items to choose from. In the context of your life, this decision is far from “existential” and therefore not very important. You use your gut/intuition/your fast brain (See Daniel Kahneman’s great book Thinking, Fast and Slow) and decide.
Who knows why you might order one thing one day and something else the next. You may not even be sure yourself. Much of your day is spent making those non “fight or flight” decisions, at work, at home, and whenever you’re interacting with the world.
Without you being aware of it, these sets of small decisions can be made into a “signature” that is unique to you. When asked about you, people may say things like “well, that person is conservative in their choices” or “when buying clothing I should be asking that person, they have the best fashion sense,” etc. This “signature” defines you as a person as much as your physical features, knowledge or abilities do.
How do you make decisions? The neural network in your brain uses its vast experience, having been exposed to years of life where events continuously impact it. The sum of your senses “teach” your brain to avoid certain things and crave others. The same process takes place in your professional life. Your accumulated experience means that you know how to approach a problem and determine a course of action without a lot of hesitation and without having to spend too much time thinking about it.
This accumulated knowledge is priceless. We all have these “hidden gems” of knowledge that are unique to us.
In the movies, artificial intelligence is often a self-aware, auto-evolving entity that craves knowledge and is able to harvest data and make sense of it by itself. Often with a British accent, for some reason. In reality, to me, a valuable AI system is one that is able to “record” and digitally store a portion of the experience of a specific human, so as to be able to mimic their decision-making process digitally.
In other words, what if we could have the ability to forecast what Albert Einstein would do given a choice of approaches to solve a problem? What if we could consult Andy Warhol on what color to choose for a school art project? Or ask Sun Tzu how to strategize our next Fortnite game?
A valuable AI system is able to focus on a certain “vertical” of a human’s experience. By using a “fly on the wall” paradigm along with a feedback loop, it should be able, in time, to have the ability to mimic that person’s decision-making process.
In the not so far future you will walk into your office craving breakfast, and your favorite deli’s delivery drone will meet you at the entrance. It will have one of those new menu items it chose for you based on your unique “digital” decision signature. While you enjoy the delicious item you will probably say to yourself “wow, how did they know exactly what I wanted?”
Going forward, this blog will be home to posts about technology, the commercial real estate industry, and other topics of interest. I promise none of the entries (after this one) will be me telling you why I think our company is great.
My background is in building software platforms in the B2B world using artificial intelligence and machine learning technologies. I’ve done this before in different industries and created technology to solve specific problems for a lot of companies.
In January of 2018, I was talking with a good friend of mine who is the founder and chairman of ABP Capital. We were discussing the commercial lending space, which is an area I knew nothing about at the time. He felt that there was a lack of products to help automate the manual activity he was putting into originating and auditing his loan portfolio. Trust me, it was a more fun conversation than it probably comes across as here. But I took his word for it, and spent several months looking into his industry.
From the point of view of a technology guy, it seemed to me like the commercial real estate lending world was far behind the residential real estate lending side. I wanted to know why. I looked at who was embracing new technology solutions on the residential side, and it was the same lenders that were in the CRE space. In other words, these institutions were willing to embrace new technology where it was available, but there did seem to be a lack of products on the commercial side.
So, I went to the best technologist I know. I reached out to Shy Blick, a business partner of mine that I had worked with before. Shy is one of the most amazing technologists in the country (you’ll hear from him on this blog). Shy felt confident that we could create something to address this issue and solve my friend’s problem. So, Shy and I brought in some terrific developers, and we started to build Blooma.
Most importantly, we didn’t just start coding. Instead, we connected with 3 commercial real estate lenders that allowed us direct access to their underwriters and analysts, so we could understand how loans were being originated and how portfolios were being audited. We mapped out their existing approaches first. We discovered that way too much of the time of a highly skilled analyst or underwriter is occupied with reading through tons of documents and third-party systems, pulling out pieces of information, and then pasting or keying them into other documents, spreadsheets, or systems of record.
We thought about what AI and machine learning could do for them in this process. We wondered if we could create something that could bundle everything together and do the data gathering, document reading, and number crunching. If we could use artificial intelligence to parse the documents and build the comp structures, we realized that this would leave the underwriters to focus on what they were really good at – the artistic side of underwriting. After all, their experience, relationships, and human decision-making ability was why they were truly valuable. What if we could do the more manual and monotonous side of their job for them and leave them the fun and interesting stuff?
Our technical team started building, but we never walked away from those underwriters and analysts. They acted as our product team, and they truly helped us create Blooma. So, in a way, we were built for CRE lenders by CRE lenders. There was something unique about the way we were able to connect these bankers who really understood the issues with a team of people who have a level of knowledge of AI, machine learning, and data analytics that very few technologists in the world have. A bunch of nerds (and I include myself there) built the technology, but it was through the output of these banks that it came to life – and I think that is ultimately the secret to why it works.
But I still needed to see more. I rarely look at a technology and just know that it is going to be huge. Ultimately, I have to hear that from the customers that will use it. We went commercial in January of 2020, and we’ve shown Blooma to a whole lot of lenders since then. The reaction is almost always genuine shock at what we’ve been able to do. Of course, most of our customers don’t really care about the technology itself, just the results – higher efficiency, driving time of underwriting way down, massively reducing the costs of originating and auditing deals. We bring them to revenue quicker because they can make decisions to lend quicker.
Blooma is out there now, and it’s fun to get the “oohs” and “ahhs” during the demos, and feedback about how great it is from our customers. But because of how it came together, I’m more convinced than ever that we need to keep listening to our customers and always continue to make improvements. After all, that’s how it all started.