Weekly Wrap: Bite-sized Automation at Wells Fargo and Ally
Automating end-to-end processes can be a big bite to chew off. This week, Bank Automation News explored best-practice conversations with leaders at Wells Fargo and Ally, who think of automation in smaller pieces — individual building blocks, like Legos, that when put together can create something larger and more significant.
BAN’s look at automation best practices began this week with a feature on mortgage automation, including a discussion with Nate Longfellow, head of digital, home lending at the $1.96 trillion Wells Fargo. In another best practices focused piece, BAN learned the $182.2 billion Ally’s three steps for turning automation into a strong, democratized discipline.
Finally, US Bank plans to use data analytics and its AI savings tool, which is powered by Personetics’ AI, to improve customer experience.
Find this and more in today’s episode of the Weekly Wrap, featuring Associate Editors Jaspreet Kalra and Loraine Lawson.
Bank Automation Ignite, on April 13-14, is the event for inspiring automation initiatives and investment in financial services. At the virtual event, financial services professionals can discover new use cases and technologies that are accelerating automation in banking. Learn more and register at www.BankAutomationIgnite.com.
Subscribe to The Buzz Podcast on iTunes, Spotify, or download the episode.
The following is a transcript generated by AI technology that has been lightly edited but still contains errors.
Loraine Lawson
Hi everyone. I’m Loraine Lawson and welcome to the Buzz from Bank Automation News, where we explore how automation technology is transforming the baking industry. This is our weekly wrap for what’s happening in industry this week. And before beginning, I’d like to give a big thanks to band sponsors and advertisers. NVIDIA and MX. Thank you so much for your support. I’m pleased to be joined by Jaspreet Kalra. It is April 2, and here are some of the biggest technology stories from this week. First of all, President Joe Biden’s massive infrastructure bill details of which were released Wednesday has earmarked 100 billion to bring affordable broadband Internet access to all Americans by 2029. The Supreme Court unanimously sided with by Facebook on Thursday ruling that its phone notification system does not violate a law prohibiting automatic telephone dialing systems. Microsoft wins the US Army contract for augmented reality headsets worth up to 21 point 9 billion over 10 years. And Amazon and Google say most of their employees will return to work this fall when some returning as early as April. So here are the news items from our editorial team this week, just spree, you drill down on mortgage automation, which some have said lags behind other banking processes when it comes to automation. And you found there were two primary buckets of automation in mortgages. Can you explain?Jaspreet Kalra
So yeah, I mean, the two primary buckets, if you look at them are probably the two categories in which automation is being carried out. The first one is processing documents, and anyone who’s gone through the mortgage process knows that it comes with about 400 pages worth of documentation. And now you have technologies like optical character recognition that can be used to read documents, extract structured data from them, and then put them into the systems where the mortgage is going to be processed. Whereas on the other hand, you have the decisioning systems or say, when you’re using artificial intelligence or machine learning to do processes like underwriting. Now, that gets a little more complicated, because more often than not these situations and require more context in the machine is currently able to absorb or assimilate. So the overall conversations that I had with sources for this story really led me to this sort of takeaway that showed us a lot of work happening. But the decisioning system were a little bit of intuition or say a little bit of a decisioning power is required. That’s where the machines still have to make progress. And that’s why we’ve seen sort of a slower uptake of automation within the mortgage verticals.
Hi everyone. I’m Loraine Lawson and welcome to the Buzz from Bank Automation News, where we explore how automation technology is transforming the baking industry. This is our weekly wrap for what’s happening in industry this week. And before beginning, I’d like to give a big thanks to band sponsors and advertisers. NVIDIA and MX. Thank you so much for your support. I’m pleased to be joined by Jaspreet Kalra. It is April 2, and here are some of the biggest technology stories from this week. First of all, President Joe Biden’s massive infrastructure bill details of which were released Wednesday has earmarked 100 billion to bring affordable broadband Internet access to all Americans by 2029. The Supreme Court unanimously sided with by Facebook on Thursday ruling that its phone notification system does not violate a law prohibiting automatic telephone dialing systems. Microsoft wins the US Army contract for augmented reality headsets worth up to 21 point 9 billion over 10 years. And Amazon and Google say most of their employees will return to work this fall when some returning as early as April. So here are the news items from our editorial team this week, just spree, you drill down on mortgage automation, which some have said lags behind other banking processes when it comes to automation. And you found there were two primary buckets of automation in mortgages. Can you explain?Jaspreet Kalra
So yeah, I mean, the two primary buckets, if you look at them are probably the two categories in which automation is being carried out. The first one is processing documents, and anyone who’s gone through the mortgage process knows that it comes with about 400 pages worth of documentation. And now you have technologies like optical character recognition that can be used to read documents, extract structured data from them, and then put them into the systems where the mortgage is going to be processed. Whereas on the other hand, you have the decisioning systems or say, when you’re using artificial intelligence or machine learning to do processes like underwriting. Now, that gets a little more complicated, because more often than not these situations and require more context in the machine is currently able to absorb or assimilate. So the overall conversations that I had with sources for this story really led me to this sort of takeaway that showed us a lot of work happening. But the decisioning system were a little bit of intuition or say a little bit of a decisioning power is required. That’s where the machines still have to make progress. And that’s why we’ve seen sort of a slower uptake of automation within the mortgage verticals.
Loraine Lawson
And you talked with Wells Fargo, which tends to be sort of on the cutting edge with automation. Where are they on automating mortgages.
And you talked with Wells Fargo, which tends to be sort of on the cutting edge with automation. Where are they on automating mortgages.
Jaspreet Kalra
So the takeaway I got from speaking with Wells Fargo was that they are very interested in leveraging data to speed up some of these processes to gain from the efficiency that comes from it to gain from this, the accuracy that comes from established data reporting procedures, but at the same time, they acknowledge that they still on the front door of using these processes, there’s a lot of say acquired intellect and knowledge that goes into underwriting a mortgage or writing a mortgage, which is why they want to take more of a building blocks are like bite sized approach to automating these processes. Because even though you have organizations like freddie mac and fannie mae already using automated underwriting systems, they’re the secondary market liquidity providers. So there, they buy mortgages in the secondary market instead of underwriting them directly. Which is why it becomes more incumbent for organizations like Wells Fargo or anyone who’s writing a mortgage directly to vet to vet the same veracity of the information that’s been put out there. So which is why Wells Fargo is interested in making sure automation delivers the efficiency gains, but they also want to do it in a step by step manner so that processes don’t break down just because you’re automating.Loraine Lawson
That makes sense. You also found out about USB as a savings tool. What is that?Jaspreet Kalra
So that was an interesting story that came out last week in the sense that US Bank is rolling out a personalized savings tool for customers in partnership with a vendor called person ethics. And what that tool essentially does is it looks at about 100 different KPIs from your consumer account, say your cash, your cash flow analysis, your balance statistics, or say, How often do you tend to spend on essential versus non essential expenditures and then recommends a saving level view. And within that app, you can also set up personalized goals to say I want to save X amount of dollars to buy a car or X amount of dollars for an education and that on the basis of those goals. The AI takes in your information about how you use your account, compares it to your to your goals and recommends a savings rate to you. So that’s really about delivering sort of advisory value to customers using machine learning.Loraine Lawson
That’s interesting. I have a tool that does something like that. It’s not associated with the bank, though it’s an independent tool. I wonder if they use AI. So I was surprised to see that an analyst Steven Greer of cylance said Adoption for that sort of tool is lagging for financial advisory platforms, did he say why?Jaspreet Kalra
So there’s a couple of reasons as to why adoption seems to have lagged for these apps, because first of all, platforms like these are solutions like these tend to target customers who are budget hawks, who are very careful about how they spend their money, how they manage their money, whereas on the other hand, you have a larger subset of customers who are more laissez faire about it, who know how much they want to save, they have a rough estimate. And that’s what they do they savings based on and the other factor is also age. So on the younger spectrum of things, you do see people now relying more on say, automated systems to recommend things to them. So sort of a headline takeaway from that conversation with Steven was also that a lot of people are now asking what can this algorithm or analytics do for me without me having to put in any effort towards it. So I think as sort of demographics, also play out might see uptick in adoption. But so far, it’s been limited. And the estimate that Stephen gave me was about seeing high single digits near about 10% has been the adoption rate for this sort of solution.Loraine Lawson
I know some banks are moving towards integrating it into their existing apps. So that may, that would, it seems that increased adoption, if it’s, you know, integrated into the app alreadyJaspreet Kalra
But moving on, I mean, I also read through one of your pieces when you talked about totok to Ally Bank, Satish Muthukrishnan, who’s the chief information data and digital officer, and you guys discuss what sort of automation Ally has been doing? What are the goals? Could you walk us through some of that?
So the takeaway I got from speaking with Wells Fargo was that they are very interested in leveraging data to speed up some of these processes to gain from the efficiency that comes from it to gain from this, the accuracy that comes from established data reporting procedures, but at the same time, they acknowledge that they still on the front door of using these processes, there’s a lot of say acquired intellect and knowledge that goes into underwriting a mortgage or writing a mortgage, which is why they want to take more of a building blocks are like bite sized approach to automating these processes. Because even though you have organizations like freddie mac and fannie mae already using automated underwriting systems, they’re the secondary market liquidity providers. So there, they buy mortgages in the secondary market instead of underwriting them directly. Which is why it becomes more incumbent for organizations like Wells Fargo or anyone who’s writing a mortgage directly to vet to vet the same veracity of the information that’s been put out there. So which is why Wells Fargo is interested in making sure automation delivers the efficiency gains, but they also want to do it in a step by step manner so that processes don’t break down just because you’re automating.Loraine Lawson
That makes sense. You also found out about USB as a savings tool. What is that?Jaspreet Kalra
So that was an interesting story that came out last week in the sense that US Bank is rolling out a personalized savings tool for customers in partnership with a vendor called person ethics. And what that tool essentially does is it looks at about 100 different KPIs from your consumer account, say your cash, your cash flow analysis, your balance statistics, or say, How often do you tend to spend on essential versus non essential expenditures and then recommends a saving level view. And within that app, you can also set up personalized goals to say I want to save X amount of dollars to buy a car or X amount of dollars for an education and that on the basis of those goals. The AI takes in your information about how you use your account, compares it to your to your goals and recommends a savings rate to you. So that’s really about delivering sort of advisory value to customers using machine learning.Loraine Lawson
That’s interesting. I have a tool that does something like that. It’s not associated with the bank, though it’s an independent tool. I wonder if they use AI. So I was surprised to see that an analyst Steven Greer of cylance said Adoption for that sort of tool is lagging for financial advisory platforms, did he say why?Jaspreet Kalra
So there’s a couple of reasons as to why adoption seems to have lagged for these apps, because first of all, platforms like these are solutions like these tend to target customers who are budget hawks, who are very careful about how they spend their money, how they manage their money, whereas on the other hand, you have a larger subset of customers who are more laissez faire about it, who know how much they want to save, they have a rough estimate. And that’s what they do they savings based on and the other factor is also age. So on the younger spectrum of things, you do see people now relying more on say, automated systems to recommend things to them. So sort of a headline takeaway from that conversation with Steven was also that a lot of people are now asking what can this algorithm or analytics do for me without me having to put in any effort towards it. So I think as sort of demographics, also play out might see uptick in adoption. But so far, it’s been limited. And the estimate that Stephen gave me was about seeing high single digits near about 10% has been the adoption rate for this sort of solution.Loraine Lawson
I know some banks are moving towards integrating it into their existing apps. So that may, that would, it seems that increased adoption, if it’s, you know, integrated into the app alreadyJaspreet Kalra
But moving on, I mean, I also read through one of your pieces when you talked about totok to Ally Bank, Satish Muthukrishnan, who’s the chief information data and digital officer, and you guys discuss what sort of automation Ally has been doing? What are the goals? Could you walk us through some of that?
Loraine Lawson
Yes, so he has a strong discipline of automation at ally. And he talked about what created that sort of culture of automation. And the first thing he said he did was to create a center of excellence for automation. And we’ve written about that before as a as a key tool, really, to make sure that you’re following best practice that you learn from previous automations, how to do better. And as a way to sort of spread automation throughout the organization, which was his second point which he has democratized automation, which I thought was an interesting way to put it. Well, it sort of manages the main platform, they have opened it up to various business teams as well. So it’s not just centralized in it, it’s democratized throughout the organization. So automation ideas can come in from all parts of ally instead of just being centrally driven by a tee. So that’s certainly a best practice that banks could take away from ally. The other is they have a tight partnership with functional and business leaders, which he said helps them identify business goals helps them identify financial cases for automation. So that was really interesting, I think, sort of those best practices that they’re following. They also are focused on three areas with automation. You talked about your buckets earlier. And it was interesting. You also talked about the bite sized automation, like the let you call them Lego pieces. That’s how he thinks about automations, Lego pieces, little building blocks. So I think that’s how we’re seeing automation actually sort of roll out is they keep one process that’s automated, and then another and then hopefully, at some point, it’ll stack and you’ll be automating across the process. But for Ally, the three areas of automation are of course RPA, deployment, workflow automation, and contextual and conversational interfaces, which basically is LS digital assistant. I started off by automating just 6% of the intense in its chatbot 6% of the time, and now it’s close to 60% containment, which means basically, when someone calls in the bot can handle the problem 60% of the time, which is pretty yeah, that’s pretty good. Um, one automation he specifically shared that I found interesting was the auto loan refinance. So whenever say you have a loan for your automobile, and you see a better rate you want to refinance. Well, before you had to collect all the documents again, and basically go through the whole process again, with the loan officer looking at it providing you with quote, you accepting and then writing a contract. Now they’ve automated that entire workflow, and what used to take 30 days now takes happens within a few hours. Okay, that was impressive.
Jaspreet Kalra
Yeah. And what did he talk about what sort of would be now I read through your piece, and it said that he so called friends did he talk about at all what’s what’s the sort of hole in one situation for them when it comes to automation?
Loraine Lawson
He did. He’s said basically, he had this very high level view of a world where you have what he called an experiment, experience economy, which is where you are able to treat every customer as if they are your only customer, or at least your primary customer. So that, you know, you come it ties into your story about personalized finance, where you come to ally and you say I’m going to create a savings account, so I can buy the dream. All right want and ally helps you set aside money for that helps you identify your savings goals, maybe breaking up into buckets. And then even to the point where at some point, it could say, Hey, your dream car has come up for sale at this dealership, and you would take your money that you’ve saved and go out and buy your dream car. So very top level, customer service experience. And of course, he says they’re not there yet. But that is the goal.
Jaspreet Kalra
Right? Makes sense. And sort of looking forward to the week that we have coming in what what are some of the stories you’re working on? What are some of the top lines that?
Loraine Lawson
Yeah, so I’m going to be working on a mobile security and bass peace next week and an in depth look at Scotia bass, automated regulation, automation, sorry, what they’re doing to regulate within their bank. But also this week, we’ll be reporting on the recent news that five federal agencies are seeking input on AI and machine learning, the Federal Reserve System, the Bureau of consumer financial protection, the fill Deposit Insurance Corporation, National Credit Union Administration, and the Office of the Comptroller of the Currency, all looking for people’s input on AI. So that should be an interesting story,
Jaspreet Kalra
certainly. And what seems to have sort of, are there were there any sort of specific things within the announcement that caught your attention or like, give us any insight into why they might be looking for this sort of advice.
Loraine Lawson
Um, as I understand it, they’re just looking for the advice because they want to see if organizations need some clarity on how to use those technologies. And they’re, they’re a little concerned about potential risks with those technologies. It’s interesting, the FCC actually did a sort of request for information in June, because it issued an advance notice of proposed rulemaking and cite comments about that time on digital technologies, including AI and Google and ml machine learning. And at that time, a wide variety of organizations and vendors responded, including the American Bankers Association, Google plaid upstart and others responded to that. So I’ve been digging through that today to find out what they had to say. So what about you? What are you reporting on for next week?
Jaspreet Kalra
So I’ve sort of been focusing on cybersecurity and ml related issues as well. So what are some of the things that I’ll be looking into the week going forward is the sort of trends automation can inspire within risk management. and risk management was really top line story this week considering what happened on wall street with certain banks like Nomura, Morgan Stanley, Goldman Sachs, etc. And the other thing I’ll be looking into is how machine learning can be attacked by something called Data poisoning in which an adversary adversarial actor or by accident, someone might end up Miss labeling or miss classifying images within it used for images or data use for training AI. So that’s the sort of ideas I’ll be looking into going going into the week forward.
Loraine Lawson
Oh, that sounds uh, that sounds mysterious, so I get data poisoning. Well, thank you so much for participating and for joining us at the weekly wrap on the buzz. For more content podcast content, check out bank automation news.com and search the buzz from Bank automation news on iTunes and Spotify. Thank you.
Automating end-to-end processes can be a big bite to chew off. This week, Bank Automation News explored best-practice conversations with leaders at Wells Fargo and Ally, who think of automation in smaller pieces — individual building blocks, like Legos, that when put together can create something larger and more significant.
BAN’s look at automation best practices began this week with a feature on mortgage automation, including a discussion with Nate Longfellow, head of digital, home lending at the $1.96 trillion Wells Fargo. In another best practices focused piece, BAN learned the $182.2 billion Ally’s three steps for turning automation into a strong, democratized discipline.
Finally, US Bank plans to use data analytics and its AI savings tool, which is powered by Personetics’ AI, to improve customer experience.
Find this and more in today’s episode of the Weekly Wrap, featuring Associate Editors Jaspreet Kalra and Loraine Lawson.
Bank Automation Ignite, on April 13-14, is the event for inspiring automation initiatives and investment in financial services. At the virtual event, financial services professionals can discover new use cases and technologies that are accelerating automation in banking. Learn more and register at www.BankAutomationIgnite.com.
Subscribe to The Buzz Podcast on iTunes, Spotify, or download the episode.
The following is a transcript generated by AI technology that has been lightly edited but still contains errors.
Loraine Lawson
Hi everyone. I’m Loraine Lawson and welcome to the Buzz from Bank Automation News, where we explore how automation technology is transforming the baking industry. This is our weekly wrap for what’s happening in industry this week. And before beginning, I’d like to give a big thanks to band sponsors and advertisers. NVIDIA and MX. Thank you so much for your support. I’m pleased to be joined by Jaspreet Kalra. It is April 2, and here are some of the biggest technology stories from this week. First of all, President Joe Biden’s massive infrastructure bill details of which were released Wednesday has earmarked 100 billion to bring affordable broadband Internet access to all Americans by 2029. The Supreme Court unanimously sided with by Facebook on Thursday ruling that its phone notification system does not violate a law prohibiting automatic telephone dialing systems. Microsoft wins the US Army contract for augmented reality headsets worth up to 21 point 9 billion over 10 years. And Amazon and Google say most of their employees will return to work this fall when some returning as early as April. So here are the news items from our editorial team this week, just spree, you drill down on mortgage automation, which some have said lags behind other banking processes when it comes to automation. And you found there were two primary buckets of automation in mortgages. Can you explain?Jaspreet Kalra
So yeah, I mean, the two primary buckets, if you look at them are probably the two categories in which automation is being carried out. The first one is processing documents, and anyone who’s gone through the mortgage process knows that it comes with about 400 pages worth of documentation. And now you have technologies like optical character recognition that can be used to read documents, extract structured data from them, and then put them into the systems where the mortgage is going to be processed. Whereas on the other hand, you have the decisioning systems or say, when you’re using artificial intelligence or machine learning to do processes like underwriting. Now, that gets a little more complicated, because more often than not these situations and require more context in the machine is currently able to absorb or assimilate. So the overall conversations that I had with sources for this story really led me to this sort of takeaway that showed us a lot of work happening. But the decisioning system were a little bit of intuition or say a little bit of a decisioning power is required. That’s where the machines still have to make progress. And that’s why we’ve seen sort of a slower uptake of automation within the mortgage verticals.
Hi everyone. I’m Loraine Lawson and welcome to the Buzz from Bank Automation News, where we explore how automation technology is transforming the baking industry. This is our weekly wrap for what’s happening in industry this week. And before beginning, I’d like to give a big thanks to band sponsors and advertisers. NVIDIA and MX. Thank you so much for your support. I’m pleased to be joined by Jaspreet Kalra. It is April 2, and here are some of the biggest technology stories from this week. First of all, President Joe Biden’s massive infrastructure bill details of which were released Wednesday has earmarked 100 billion to bring affordable broadband Internet access to all Americans by 2029. The Supreme Court unanimously sided with by Facebook on Thursday ruling that its phone notification system does not violate a law prohibiting automatic telephone dialing systems. Microsoft wins the US Army contract for augmented reality headsets worth up to 21 point 9 billion over 10 years. And Amazon and Google say most of their employees will return to work this fall when some returning as early as April. So here are the news items from our editorial team this week, just spree, you drill down on mortgage automation, which some have said lags behind other banking processes when it comes to automation. And you found there were two primary buckets of automation in mortgages. Can you explain?Jaspreet Kalra
So yeah, I mean, the two primary buckets, if you look at them are probably the two categories in which automation is being carried out. The first one is processing documents, and anyone who’s gone through the mortgage process knows that it comes with about 400 pages worth of documentation. And now you have technologies like optical character recognition that can be used to read documents, extract structured data from them, and then put them into the systems where the mortgage is going to be processed. Whereas on the other hand, you have the decisioning systems or say, when you’re using artificial intelligence or machine learning to do processes like underwriting. Now, that gets a little more complicated, because more often than not these situations and require more context in the machine is currently able to absorb or assimilate. So the overall conversations that I had with sources for this story really led me to this sort of takeaway that showed us a lot of work happening. But the decisioning system were a little bit of intuition or say a little bit of a decisioning power is required. That’s where the machines still have to make progress. And that’s why we’ve seen sort of a slower uptake of automation within the mortgage verticals.
Loraine Lawson
And you talked with Wells Fargo, which tends to be sort of on the cutting edge with automation. Where are they on automating mortgages.
And you talked with Wells Fargo, which tends to be sort of on the cutting edge with automation. Where are they on automating mortgages.
Jaspreet Kalra
So the takeaway I got from speaking with Wells Fargo was that they are very interested in leveraging data to speed up some of these processes to gain from the efficiency that comes from it to gain from this, the accuracy that comes from established data reporting procedures, but at the same time, they acknowledge that they still on the front door of using these processes, there’s a lot of say acquired intellect and knowledge that goes into underwriting a mortgage or writing a mortgage, which is why they want to take more of a building blocks are like bite sized approach to automating these processes. Because even though you have organizations like freddie mac and fannie mae already using automated underwriting systems, they’re the secondary market liquidity providers. So there, they buy mortgages in the secondary market instead of underwriting them directly. Which is why it becomes more incumbent for organizations like Wells Fargo or anyone who’s writing a mortgage directly to vet to vet the same veracity of the information that’s been put out there. So which is why Wells Fargo is interested in making sure automation delivers the efficiency gains, but they also want to do it in a step by step manner so that processes don’t break down just because you’re automating.Loraine Lawson
That makes sense. You also found out about USB as a savings tool. What is that?Jaspreet Kalra
So that was an interesting story that came out last week in the sense that US Bank is rolling out a personalized savings tool for customers in partnership with a vendor called person ethics. And what that tool essentially does is it looks at about 100 different KPIs from your consumer account, say your cash, your cash flow analysis, your balance statistics, or say, How often do you tend to spend on essential versus non essential expenditures and then recommends a saving level view. And within that app, you can also set up personalized goals to say I want to save X amount of dollars to buy a car or X amount of dollars for an education and that on the basis of those goals. The AI takes in your information about how you use your account, compares it to your to your goals and recommends a savings rate to you. So that’s really about delivering sort of advisory value to customers using machine learning.Loraine Lawson
That’s interesting. I have a tool that does something like that. It’s not associated with the bank, though it’s an independent tool. I wonder if they use AI. So I was surprised to see that an analyst Steven Greer of cylance said Adoption for that sort of tool is lagging for financial advisory platforms, did he say why?Jaspreet Kalra
So there’s a couple of reasons as to why adoption seems to have lagged for these apps, because first of all, platforms like these are solutions like these tend to target customers who are budget hawks, who are very careful about how they spend their money, how they manage their money, whereas on the other hand, you have a larger subset of customers who are more laissez faire about it, who know how much they want to save, they have a rough estimate. And that’s what they do they savings based on and the other factor is also age. So on the younger spectrum of things, you do see people now relying more on say, automated systems to recommend things to them. So sort of a headline takeaway from that conversation with Steven was also that a lot of people are now asking what can this algorithm or analytics do for me without me having to put in any effort towards it. So I think as sort of demographics, also play out might see uptick in adoption. But so far, it’s been limited. And the estimate that Stephen gave me was about seeing high single digits near about 10% has been the adoption rate for this sort of solution.Loraine Lawson
I know some banks are moving towards integrating it into their existing apps. So that may, that would, it seems that increased adoption, if it’s, you know, integrated into the app alreadyJaspreet Kalra
But moving on, I mean, I also read through one of your pieces when you talked about totok to Ally Bank, Satish Muthukrishnan, who’s the chief information data and digital officer, and you guys discuss what sort of automation Ally has been doing? What are the goals? Could you walk us through some of that?
So the takeaway I got from speaking with Wells Fargo was that they are very interested in leveraging data to speed up some of these processes to gain from the efficiency that comes from it to gain from this, the accuracy that comes from established data reporting procedures, but at the same time, they acknowledge that they still on the front door of using these processes, there’s a lot of say acquired intellect and knowledge that goes into underwriting a mortgage or writing a mortgage, which is why they want to take more of a building blocks are like bite sized approach to automating these processes. Because even though you have organizations like freddie mac and fannie mae already using automated underwriting systems, they’re the secondary market liquidity providers. So there, they buy mortgages in the secondary market instead of underwriting them directly. Which is why it becomes more incumbent for organizations like Wells Fargo or anyone who’s writing a mortgage directly to vet to vet the same veracity of the information that’s been put out there. So which is why Wells Fargo is interested in making sure automation delivers the efficiency gains, but they also want to do it in a step by step manner so that processes don’t break down just because you’re automating.Loraine Lawson
That makes sense. You also found out about USB as a savings tool. What is that?Jaspreet Kalra
So that was an interesting story that came out last week in the sense that US Bank is rolling out a personalized savings tool for customers in partnership with a vendor called person ethics. And what that tool essentially does is it looks at about 100 different KPIs from your consumer account, say your cash, your cash flow analysis, your balance statistics, or say, How often do you tend to spend on essential versus non essential expenditures and then recommends a saving level view. And within that app, you can also set up personalized goals to say I want to save X amount of dollars to buy a car or X amount of dollars for an education and that on the basis of those goals. The AI takes in your information about how you use your account, compares it to your to your goals and recommends a savings rate to you. So that’s really about delivering sort of advisory value to customers using machine learning.Loraine Lawson
That’s interesting. I have a tool that does something like that. It’s not associated with the bank, though it’s an independent tool. I wonder if they use AI. So I was surprised to see that an analyst Steven Greer of cylance said Adoption for that sort of tool is lagging for financial advisory platforms, did he say why?Jaspreet Kalra
So there’s a couple of reasons as to why adoption seems to have lagged for these apps, because first of all, platforms like these are solutions like these tend to target customers who are budget hawks, who are very careful about how they spend their money, how they manage their money, whereas on the other hand, you have a larger subset of customers who are more laissez faire about it, who know how much they want to save, they have a rough estimate. And that’s what they do they savings based on and the other factor is also age. So on the younger spectrum of things, you do see people now relying more on say, automated systems to recommend things to them. So sort of a headline takeaway from that conversation with Steven was also that a lot of people are now asking what can this algorithm or analytics do for me without me having to put in any effort towards it. So I think as sort of demographics, also play out might see uptick in adoption. But so far, it’s been limited. And the estimate that Stephen gave me was about seeing high single digits near about 10% has been the adoption rate for this sort of solution.Loraine Lawson
I know some banks are moving towards integrating it into their existing apps. So that may, that would, it seems that increased adoption, if it’s, you know, integrated into the app alreadyJaspreet Kalra
But moving on, I mean, I also read through one of your pieces when you talked about totok to Ally Bank, Satish Muthukrishnan, who’s the chief information data and digital officer, and you guys discuss what sort of automation Ally has been doing? What are the goals? Could you walk us through some of that?
Loraine Lawson
Yes, so he has a strong discipline of automation at ally. And he talked about what created that sort of culture of automation. And the first thing he said he did was to create a center of excellence for automation. And we’ve written about that before as a as a key tool, really, to make sure that you’re following best practice that you learn from previous automations, how to do better. And as a way to sort of spread automation throughout the organization, which was his second point which he has democratized automation, which I thought was an interesting way to put it. Well, it sort of manages the main platform, they have opened it up to various business teams as well. So it’s not just centralized in it, it’s democratized throughout the organization. So automation ideas can come in from all parts of ally instead of just being centrally driven by a tee. So that’s certainly a best practice that banks could take away from ally. The other is they have a tight partnership with functional and business leaders, which he said helps them identify business goals helps them identify financial cases for automation. So that was really interesting, I think, sort of those best practices that they’re following. They also are focused on three areas with automation. You talked about your buckets earlier. And it was interesting. You also talked about the bite sized automation, like the let you call them Lego pieces. That’s how he thinks about automations, Lego pieces, little building blocks. So I think that’s how we’re seeing automation actually sort of roll out is they keep one process that’s automated, and then another and then hopefully, at some point, it’ll stack and you’ll be automating across the process. But for Ally, the three areas of automation are of course RPA, deployment, workflow automation, and contextual and conversational interfaces, which basically is LS digital assistant. I started off by automating just 6% of the intense in its chatbot 6% of the time, and now it’s close to 60% containment, which means basically, when someone calls in the bot can handle the problem 60% of the time, which is pretty yeah, that’s pretty good. Um, one automation he specifically shared that I found interesting was the auto loan refinance. So whenever say you have a loan for your automobile, and you see a better rate you want to refinance. Well, before you had to collect all the documents again, and basically go through the whole process again, with the loan officer looking at it providing you with quote, you accepting and then writing a contract. Now they’ve automated that entire workflow, and what used to take 30 days now takes happens within a few hours. Okay, that was impressive.
Jaspreet Kalra
Yeah. And what did he talk about what sort of would be now I read through your piece, and it said that he so called friends did he talk about at all what’s what’s the sort of hole in one situation for them when it comes to automation?
Loraine Lawson
He did. He’s said basically, he had this very high level view of a world where you have what he called an experiment, experience economy, which is where you are able to treat every customer as if they are your only customer, or at least your primary customer. So that, you know, you come it ties into your story about personalized finance, where you come to ally and you say I’m going to create a savings account, so I can buy the dream. All right want and ally helps you set aside money for that helps you identify your savings goals, maybe breaking up into buckets. And then even to the point where at some point, it could say, Hey, your dream car has come up for sale at this dealership, and you would take your money that you’ve saved and go out and buy your dream car. So very top level, customer service experience. And of course, he says they’re not there yet. But that is the goal.
Jaspreet Kalra
Right? Makes sense. And sort of looking forward to the week that we have coming in what what are some of the stories you’re working on? What are some of the top lines that?
Loraine Lawson
Yeah, so I’m going to be working on a mobile security and bass peace next week and an in depth look at Scotia bass, automated regulation, automation, sorry, what they’re doing to regulate within their bank. But also this week, we’ll be reporting on the recent news that five federal agencies are seeking input on AI and machine learning, the Federal Reserve System, the Bureau of consumer financial protection, the fill Deposit Insurance Corporation, National Credit Union Administration, and the Office of the Comptroller of the Currency, all looking for people’s input on AI. So that should be an interesting story,
Jaspreet Kalra
certainly. And what seems to have sort of, are there were there any sort of specific things within the announcement that caught your attention or like, give us any insight into why they might be looking for this sort of advice.
Loraine Lawson
Um, as I understand it, they’re just looking for the advice because they want to see if organizations need some clarity on how to use those technologies. And they’re, they’re a little concerned about potential risks with those technologies. It’s interesting, the FCC actually did a sort of request for information in June, because it issued an advance notice of proposed rulemaking and cite comments about that time on digital technologies, including AI and Google and ml machine learning. And at that time, a wide variety of organizations and vendors responded, including the American Bankers Association, Google plaid upstart and others responded to that. So I’ve been digging through that today to find out what they had to say. So what about you? What are you reporting on for next week?
Jaspreet Kalra
So I’ve sort of been focusing on cybersecurity and ml related issues as well. So what are some of the things that I’ll be looking into the week going forward is the sort of trends automation can inspire within risk management. and risk management was really top line story this week considering what happened on wall street with certain banks like Nomura, Morgan Stanley, Goldman Sachs, etc. And the other thing I’ll be looking into is how machine learning can be attacked by something called Data poisoning in which an adversary adversarial actor or by accident, someone might end up Miss labeling or miss classifying images within it used for images or data use for training AI. So that’s the sort of ideas I’ll be looking into going going into the week forward.
Loraine Lawson
Oh, that sounds uh, that sounds mysterious, so I get data poisoning. Well, thank you so much for participating and for joining us at the weekly wrap on the buzz. For more content podcast content, check out bank automation news.com and search the buzz from Bank automation news on iTunes and Spotify. Thank you.
