JJ Hornblass
Hi, everyone. I’m JJ Hornblass and welcome to the Buzz from Bank Automation News where we chart the future of banking technology and automation. We welcome you to this weekly wrap for what’s happening in banking automation on May 21 2021. Before beginning, we thank our advertisers MX and Fenergo for their support, so thank you to them, and pleased to be joined by Loraine Lawson and Myra Thomas of the Bank Automation News team. Hello to both of you. First, we’ll get to some general technology news. A crypto currency crackdown seems to be in the works. Yesterday, US Federal Reserve Chief Jerome Powell said that cryptocurrency posed a risk to financial stability hinting at possible regulation down the road. Also yesterday the Treasury Department said that it would require transfers of more cryptocurrency transfers of more than $10,000 to be reported to the Internal Revenue Service. carmakers are getting no special treatment when it comes to the semiconductor chip shortage. Yesterday, the Commerce Secretary Gina Raimondo told the Wall Street Journal that there are no easy solutions to the chip chip shortage that has caused the American auto industry to halt assembly lines. at Apple lawyers are planning to put Chief Executive Officer Tim Cook on the witness stand today to help drive home its defense against Epic Games. antitrust claims against Apple epic filed its lawsuit against Apple last August after the iPhone maker kicked fortnight which is an epic game product out of the iTunes App Store for violating the app store’s rules to banking, automation, industry news, we’re going to focus on two topics today. Starting with methods for calculating bad KPIs, there has been some new insights into how to view bought KPIs and and calculate them. And Lorraine, why don’t you start by giving us some background on just maybe KPI calculations in general in financial services, and then we can move to the more complex body calculations around KPIs. Yes.Loraine Lawson
Key Performance Indicators. One of the things to know is that traditionally with budgets people have measured them in very operational terms. How many bots Have you deployed? What what are the bots doing? So there’s a sort of very bot focus KPIs. But now what they’re saying is that bots should be API should be based on what the business goal is. So looking more at what are your goals for measuring, so what are your goals for the but in terms of the business? So for instance, you would want to measure a process before you automate it. So you can know what the automation did for you. Did it reduce error rates, for instance, to know, if it mattered, you have to know what the error rates were before? Right? So they’re advising people to engage in process discovery and process mining to sort of identify KPIs before the bots are deployed? And to really look at it more from what are you trying to achieve? Is it going to help you earn more money? For instance, is it going to help you turn around loans faster? These are the kind of KPIs they’re recommending that people pursue.JJ Hornblass
Is there a sense for whether or not the KPIs whether or not the KPIs are material enough? He ran through in your in an article you wrote this week, you ran through some of the metrics, you know, the possible metrics for for one particular bank, just as an example. Are these numbers significant enough to allocate resources to bot development?Loraine Lawson
Well, that’s one thing you Have to define and the only way you can do that is if you know going in what the process was before you started automating it, right? So you should, for instance, know, okay, so let’s say Happy state bank, which is based in Texas, I’m really Texas, they knew it took them six to eight hours, on average to build the bot, from defining the process to testing the automation, they had measured that. So what they found was when they deployed the bots, that it the 40 bucks saved roughly 10,000 hours of employee time per year. So obviously, in that case, it’s worth it because he traded 62 hours for 10,000 hours. Well, that’s

JJ Hornblass
Yeah, assuming that the price for each of those hours is equal. And that they that they don’t have some upfront costs to to a vendor for for the bots. Is there is there a sense for where I understand that it kind of boils down to hours, but even thinking about segments? Is there segments of banking is there? Are there segments of banking where the time sink is just particularly acute, and therefore the KPI calculus comes out? Again, generally with the with the, with the with the caveats you’ve presented, where it generally comes out in the in the financial institutions favor?

Loraine Lawson
Favorite, or favorite to play the but? Well, I think that’s what they’re talking about business value, you have to look beyond just full time equivalency, like you said, All employees aren’t equivalent. So like developers cost more money. So while that’s one way to look at it, you should also be looking at what business value you’re getting from that. But is it lowering costs? is increasing sales? is driving better financial performance? Or in the pace place of compliance? Is it reducing errors, which bots tend to do?

JJ Hornblass
Right? My read Lorraine mentioned process discovery as part of the KPI calculus. I know that that’s k process discovery generally, is becoming more of a focus in the sector. What have you learned about process discovery from maybe particular from the from the interests of particular parties in the industry?

Myra Thomas
Yeah, I read an interview with Matt Carbonara from Citi ventures. And he specifically referenced the firm’s interest in investing in more process discovery organizations, and specifically, because of the dilemmas that banks are facing in order to, you know, get to that KPI, you know, that they want to find out that that our return on investment that they ultimately need, there’s a problem in the initial collection of what you’re trying to actually automate. And so they recently invested in an organization called scan. And, you know, he sort of ran it down because he talked about the process discovery process. And this AI enabled technology, and this firm that they invested in, actually removes the human touch from trying to figure out what sort of processes you’re trying to automate, which actually ends up becoming a big problem. And looking at the KPI. So, you know, basically, with this sort of AI enabled technology, which is not just a scam thing, you can actually use sort of computer vision, you know, it combines data science along with computer vision that actually looks at all the digital traces of the individual as they work, whether it’s across, you know, underwriting or, you know, customer onboarding, or servicing. So it makes the whole process, you know, much more seamless, and better to understand what you’re working on, in order to figure out how to process you know, automate it well, and then figure out, you know, sort of the qualitative measures that are needed to understand whether or not this automation has worked, because we’re trying to move away from more of a quantitative like, so if you’re looking at business processes versus just money, we need to really understand, like, I think the rains mentioned in her piece, how do you actually get to, you know, the business goal, and not just talk about money. I mean, obviously, money is important too. But, you know, there’s got to be some sort of measure of what you’re actually looking to automate and how that relates to your ultimate strategy.

JJ Hornblass
So maybe Lorraine, maybe the rain process, sort of what considering considering topic of the day process what Myra has just said and kind of translated into the KPI mindset. Where does where does, you know, is this visualization? You know, how would you KPI that like, where would that feed into a KPI formula?

Loraine Lawson
Where we’re visualizing the process?

JJ Hornblass
Well, the visualization allows you to understand the hours contributed for particular process. Correct. So, you know, I guess all your your ending up doing there is just calculating the hours contributed to the particular process. But you’ve got a methodology as Myra spelled out, you’ve got a methodology for identifying that. Does that make sense? Or would you think about it in a different way?

Loraine Lawson
Well, I think you know, what this speaks to is you’re you’re looking beyond Well, how Who am I? How many people can I get rid of with this to looking at what that process is actually trying to achieve in the business? Right. So that’s, that’s how I see it is trying to understand well, the process, what are you doing with the process, that’s what the bot should be achieving? That’s the goal the bot should be achieving. And I will say that one of the reasons we’re talking about this, I think, is that a lot of vendors are adding process discovery to there. That’s a relatively new innovation to to add process discovery and process mining to their products, which helps banks better be able to identify early on their automation goals. Lorraine,

Myra Thomas
can you explain briefly, I guess, because I think there’s some, you know, misunderstanding between process mining and process discovery.

Loraine Lawson
Fair. So process mining is sort of our umbrella term for looking at the process, determining what it is process discovery is more of a subset of that that looks at what is the process before we automate, what are you actually doing, it discovers not just what the employee is doing, but what the systems are doing. So there’s very much an automated aspect to it, and that it’s looking at Well, here’s what the system does in which system it interacts with when you push this button. So looking at system records, as part of process discovery, so that you can do better process mining

Myra Thomas
it because I think ultimately, I ever forget, I think it was someone from CBN sites that I was speaking with, who talked about, you know, bots are only as good as the data that they get. And so data always, you know, is ultimately touched by people to some respects. And so if you can, as much as you can remove people from the equation, I think, yeah, that makes sense. And I think that’s what process discovery is really about.

Loraine Lawson
Yes, it’s that Yeah. But as I say, it’s not just looking at the people it’s looking with the systems do, because that’s not always clearly defined. I mean, you think we talk about a process to sort of understand it, these shapes have, you know, hundreds 1000s of processes that are running when you consider their information systems as part of that. So they don’t always know exactly what’s happening where, when, and that’s part of what the process discovery automation does.

JJ Hornblass
Is there is there one kind of particular example of a vendor adding process, discovery features, that that kind of comes to mind? I know

Loraine Lawson
Kryon has done it, because they they’ve been talking a lot about process discovery, and like their webinars and sort of building a center of excellence around that with their product. And I know of course, UiPath as that aspect. They talk a lot about it. I would say most of the major vendors at this point have process discovery.

JJ Hornblass
Very good. Well, very interesting. What do we have on tap for next week on bank automation? news.com?

Myra Thomas
Absolutely. We have a feature coming up that looks at the implications of y bore as far as processes for banks that really delves in depth into into that issue, given that banks are looking to comply by the end of this year. We’re also looking at I think, Lorraine, you have a great piece for Center of Excellence section of the publication.

Loraine Lawson
Yes, we’ll be looking at there two things that will probably come up next week. That will come up next week. One is best practices for bots. So some of the same people that talk to you for KPS, plus some additional sources looking at what are your best practices for deploying bots so that you get the most value out of them. And then also, we maybe you’ve mentioned data earlier, and how People touch the data. Well, we’re going to be looking at how the data can poisoning your AI to be biased. And so looking at sort of how banks can be ethical with their AI and their data.

Myra Thomas
Yeah, particularly looking at credit underwriting decisions. Also, we’re hoping in the future to provide a little bit more information on the implications of some of the big banks offering cryptocurrency, for their wealthy clients, and what that means for the implications for automation, particularly with what you said, JJ, as far as regulators sort of getting interested in sort of going after some of these organizations, which makes it kind of interesting, given the fact that we’ve got wells, I think, trying to think of who else B and y Mellon, and a number of the really big banks have recently entered into the wealth management space with cryptocurrency.

JJ Hornblass
Yeah. Sounds good. Thank you to both to both of you, excuse me for, for being with us today and sharing these developments. And thank you to all of our listeners for tuning in. We look forward to seeing you on future episodes of the buzz. Meanwhile, please don’t hesitate to follow us on Twitter and LinkedIn. And of course, visit us at Bank automation news.com we’ll see you next time.

JJ Hornblass
Hi, everyone. I’m JJ Hornblass and welcome to the Buzz from Bank Automation News where we chart the future of banking technology and automation. We welcome you to this weekly wrap for what’s happening in banking automation on May 21 2021. Before beginning, we thank our advertisers MX and Fenergo for their support, so thank you to them, and pleased to be joined by Loraine Lawson and Myra Thomas of the Bank Automation News team. Hello to both of you. First, we’ll get to some general technology news. A crypto currency crackdown seems to be in the works. Yesterday, US Federal Reserve Chief Jerome Powell said that cryptocurrency posed a risk to financial stability hinting at possible regulation down the road. Also yesterday the Treasury Department said that it would require transfers of more cryptocurrency transfers of more than $10,000 to be reported to the Internal Revenue Service. carmakers are getting no special treatment when it comes to the semiconductor chip shortage. Yesterday, the Commerce Secretary Gina Raimondo told the Wall Street Journal that there are no easy solutions to the chip chip shortage that has caused the American auto industry to halt assembly lines. at Apple lawyers are planning to put Chief Executive Officer Tim Cook on the witness stand today to help drive home its defense against Epic Games. antitrust claims against Apple epic filed its lawsuit against Apple last August after the iPhone maker kicked fortnight which is an epic game product out of the iTunes App Store for violating the app store’s rules to banking, automation, industry news, we’re going to focus on two topics today. Starting with methods for calculating bad KPIs, there has been some new insights into how to view bought KPIs and and calculate them. And Lorraine, why don’t you start by giving us some background on just maybe KPI calculations in general in financial services, and then we can move to the more complex body calculations around KPIs. Yes.Loraine Lawson
Key Performance Indicators. One of the things to know is that traditionally with budgets people have measured them in very operational terms. How many bots Have you deployed? What what are the bots doing? So there’s a sort of very bot focus KPIs. But now what they’re saying is that bots should be API should be based on what the business goal is. So looking more at what are your goals for measuring, so what are your goals for the but in terms of the business? So for instance, you would want to measure a process before you automate it. So you can know what the automation did for you. Did it reduce error rates, for instance, to know, if it mattered, you have to know what the error rates were before? Right? So they’re advising people to engage in process discovery and process mining to sort of identify KPIs before the bots are deployed? And to really look at it more from what are you trying to achieve? Is it going to help you earn more money? For instance, is it going to help you turn around loans faster? These are the kind of KPIs they’re recommending that people pursue.JJ Hornblass
Is there a sense for whether or not the KPIs whether or not the KPIs are material enough? He ran through in your in an article you wrote this week, you ran through some of the metrics, you know, the possible metrics for for one particular bank, just as an example. Are these numbers significant enough to allocate resources to bot development?Loraine Lawson
Well, that’s one thing you Have to define and the only way you can do that is if you know going in what the process was before you started automating it, right? So you should, for instance, know, okay, so let’s say Happy state bank, which is based in Texas, I’m really Texas, they knew it took them six to eight hours, on average to build the bot, from defining the process to testing the automation, they had measured that. So what they found was when they deployed the bots, that it the 40 bucks saved roughly 10,000 hours of employee time per year. So obviously, in that case, it’s worth it because he traded 62 hours for 10,000 hours. Well, that’s

JJ Hornblass
Yeah, assuming that the price for each of those hours is equal. And that they that they don’t have some upfront costs to to a vendor for for the bots. Is there is there a sense for where I understand that it kind of boils down to hours, but even thinking about segments? Is there segments of banking is there? Are there segments of banking where the time sink is just particularly acute, and therefore the KPI calculus comes out? Again, generally with the with the, with the with the caveats you’ve presented, where it generally comes out in the in the financial institutions favor?

Loraine Lawson
Favorite, or favorite to play the but? Well, I think that’s what they’re talking about business value, you have to look beyond just full time equivalency, like you said, All employees aren’t equivalent. So like developers cost more money. So while that’s one way to look at it, you should also be looking at what business value you’re getting from that. But is it lowering costs? is increasing sales? is driving better financial performance? Or in the pace place of compliance? Is it reducing errors, which bots tend to do?

JJ Hornblass
Right? My read Lorraine mentioned process discovery as part of the KPI calculus. I know that that’s k process discovery generally, is becoming more of a focus in the sector. What have you learned about process discovery from maybe particular from the from the interests of particular parties in the industry?

Myra Thomas
Yeah, I read an interview with Matt Carbonara from Citi ventures. And he specifically referenced the firm’s interest in investing in more process discovery organizations, and specifically, because of the dilemmas that banks are facing in order to, you know, get to that KPI, you know, that they want to find out that that our return on investment that they ultimately need, there’s a problem in the initial collection of what you’re trying to actually automate. And so they recently invested in an organization called scan. And, you know, he sort of ran it down because he talked about the process discovery process. And this AI enabled technology, and this firm that they invested in, actually removes the human touch from trying to figure out what sort of processes you’re trying to automate, which actually ends up becoming a big problem. And looking at the KPI. So, you know, basically, with this sort of AI enabled technology, which is not just a scam thing, you can actually use sort of computer vision, you know, it combines data science along with computer vision that actually looks at all the digital traces of the individual as they work, whether it’s across, you know, underwriting or, you know, customer onboarding, or servicing. So it makes the whole process, you know, much more seamless, and better to understand what you’re working on, in order to figure out how to process you know, automate it well, and then figure out, you know, sort of the qualitative measures that are needed to understand whether or not this automation has worked, because we’re trying to move away from more of a quantitative like, so if you’re looking at business processes versus just money, we need to really understand, like, I think the rains mentioned in her piece, how do you actually get to, you know, the business goal, and not just talk about money. I mean, obviously, money is important too. But, you know, there’s got to be some sort of measure of what you’re actually looking to automate and how that relates to your ultimate strategy.

JJ Hornblass
So maybe Lorraine, maybe the rain process, sort of what considering considering topic of the day process what Myra has just said and kind of translated into the KPI mindset. Where does where does, you know, is this visualization? You know, how would you KPI that like, where would that feed into a KPI formula?

Loraine Lawson
Where we’re visualizing the process?

JJ Hornblass
Well, the visualization allows you to understand the hours contributed for particular process. Correct. So, you know, I guess all your your ending up doing there is just calculating the hours contributed to the particular process. But you’ve got a methodology as Myra spelled out, you’ve got a methodology for identifying that. Does that make sense? Or would you think about it in a different way?

Loraine Lawson
Well, I think you know, what this speaks to is you’re you’re looking beyond Well, how Who am I? How many people can I get rid of with this to looking at what that process is actually trying to achieve in the business? Right. So that’s, that’s how I see it is trying to understand well, the process, what are you doing with the process, that’s what the bot should be achieving? That’s the goal the bot should be achieving. And I will say that one of the reasons we’re talking about this, I think, is that a lot of vendors are adding process discovery to there. That’s a relatively new innovation to to add process discovery and process mining to their products, which helps banks better be able to identify early on their automation goals. Lorraine,

Myra Thomas
can you explain briefly, I guess, because I think there’s some, you know, misunderstanding between process mining and process discovery.

Loraine Lawson
Fair. So process mining is sort of our umbrella term for looking at the process, determining what it is process discovery is more of a subset of that that looks at what is the process before we automate, what are you actually doing, it discovers not just what the employee is doing, but what the systems are doing. So there’s very much an automated aspect to it, and that it’s looking at Well, here’s what the system does in which system it interacts with when you push this button. So looking at system records, as part of process discovery, so that you can do better process mining

Myra Thomas
it because I think ultimately, I ever forget, I think it was someone from CBN sites that I was speaking with, who talked about, you know, bots are only as good as the data that they get. And so data always, you know, is ultimately touched by people to some respects. And so if you can, as much as you can remove people from the equation, I think, yeah, that makes sense. And I think that’s what process discovery is really about.

Loraine Lawson
Yes, it’s that Yeah. But as I say, it’s not just looking at the people it’s looking with the systems do, because that’s not always clearly defined. I mean, you think we talk about a process to sort of understand it, these shapes have, you know, hundreds 1000s of processes that are running when you consider their information systems as part of that. So they don’t always know exactly what’s happening where, when, and that’s part of what the process discovery automation does.

JJ Hornblass
Is there is there one kind of particular example of a vendor adding process, discovery features, that that kind of comes to mind? I know

Loraine Lawson
Kryon has done it, because they they’ve been talking a lot about process discovery, and like their webinars and sort of building a center of excellence around that with their product. And I know of course, UiPath as that aspect. They talk a lot about it. I would say most of the major vendors at this point have process discovery.

JJ Hornblass
Very good. Well, very interesting. What do we have on tap for next week on bank automation? news.com?

Myra Thomas
Absolutely. We have a feature coming up that looks at the implications of y bore as far as processes for banks that really delves in depth into into that issue, given that banks are looking to comply by the end of this year. We’re also looking at I think, Lorraine, you have a great piece for Center of Excellence section of the publication.

Loraine Lawson
Yes, we’ll be looking at there two things that will probably come up next week. That will come up next week. One is best practices for bots. So some of the same people that talk to you for KPS, plus some additional sources looking at what are your best practices for deploying bots so that you get the most value out of them. And then also, we maybe you’ve mentioned data earlier, and how People touch the data. Well, we’re going to be looking at how the data can poisoning your AI to be biased. And so looking at sort of how banks can be ethical with their AI and their data.

Myra Thomas
Yeah, particularly looking at credit underwriting decisions. Also, we’re hoping in the future to provide a little bit more information on the implications of some of the big banks offering cryptocurrency, for their wealthy clients, and what that means for the implications for automation, particularly with what you said, JJ, as far as regulators sort of getting interested in sort of going after some of these organizations, which makes it kind of interesting, given the fact that we’ve got wells, I think, trying to think of who else B and y Mellon, and a number of the really big banks have recently entered into the wealth management space with cryptocurrency.

JJ Hornblass
Yeah. Sounds good. Thank you to both to both of you, excuse me for, for being with us today and sharing these developments. And thank you to all of our listeners for tuning in. We look forward to seeing you on future episodes of the buzz. Meanwhile, please don’t hesitate to follow us on Twitter and LinkedIn. And of course, visit us at Bank automation news.com we’ll see you next time.