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.

This is our weekly wrap for what’s happening in the industry on May 22, 2021. Before we begin a thanks to our advertisers, MX and CI&T for their support. So thank you to them.

And I’m pleased to be joined by Mary Thomas, Loraine Lawson and Jaspreet Kalra of the bank automation news team. Good to see you all.

First to some general technology news. CNBC, excuse me is reporting that Swedish payments startup klarna is close to raising a new round of venture funding at a valuation of approximately 40 billion. And this would be after raising a billion dollars of venture funding at a $31 billion valuation. Just last March. savings and investment app acorns is merging with SPAC to go public in a deal that would value acorns at roughly 2.2 billion more than double its last private valuation and hallows which lets financial institutions provide digital asset trading services raised $40 million in the series A that was led by a 16 z otherwise known as Andreessen Horowitz, and pay pal ventures among other investors to the banking industry, banking automation industry. news this week, several of the major CEOs or the CEOs of major financial institutions, testified in front of a US Senate committee to talk about the banking industry. There were several points that related to technology, automation and AI applications. jasprit. Can you give us a rundown of what was said during this this Senate hearing?

Jaspreet Kalra
we can divide what was said in two distinct buckets. One was written testimonies that came in from the CEOs right before the hearing started. And then the meat of the hearings themselves. So when it comes to the meat of the hearings themselves, a lot of the tech related questions focused on cybersecurity, which was the natural leader given that the US colonial pipeline hack or the colonial pipeline ransomware attack really played on the minds of the US senators while they were talking to banks, and they had very pointed questions on what banks specifically were doing to protect the sort of infrastructure that enables financial activities in the US. When it comes to their written testimonies. Most of them made a note of how artificial intelligence can help prevent payments, fraud in banking can help guard against malicious activity. And also the fact that Wells Fargo mentioned AI could be useful for furthering financial inclusion sort of demonstrates that the big banks are growing their appetite for leveraging this technology. But at the same time, they want to be very cautious about the ethical implications of it. And Sidney banks CEO Jean Frazier made a special note of this, in her written testimony noting that said he was one of the first companies to develop their own ethical AI framework, and would have to dig a little deeper into what they mean by the ethical AI framework. But the overall crux of it seems that AI is going to continue to seep in to the way the financial institutions guard against fraud, think about financial inclusion, and to an extent also think about how AI powered credit can further those ideas.

JJ Hornblass
Was there a sense that the senators were buying this the that these that the Well, first of all, we should also say first, you know, which which financial institutions testified before the Senate Banking, they just want to make sure we have that clear.

Jaspreet Kalra
So it was Wells Fargo, JP Morgan Chase, Bank of America, Citibank, Goldman Sachs and Morgan Stanley.

JJ Hornblass
Okay. So coming back to the question, so it did. Well, first of all, I you know, much of what you’re citing was in written testimony. Did the did the committee raise these issues that was in did they discuss, excuse me, these issues that were raised in the written testimony,

Jaspreet Kalra
the two issues that senators raised specifically was cyber security and also how the cryptocurrency space is growing. And whether the associated risks could spill into spillover into the financial ecosystem. And another point that they made an extreme note of was how cryptocurrencies are enabling some of these ransomware attacks. And a couple of senators also noted that they’d be willing to cooperate further with the banks on how to sort of fight this menace. But that was an ambiguous part, considering that the banks themselves don’t control the flow of these cryptocurrency payments, they usually happen through either exchanges, or private sort of decentralized exchanges that aren’t really owned by anyone is just community operate. So yeah, go ahead. Sorry. So mean, yeah, all I’m saying there is that the senators were certainly cautioned and alarmed at what’s going on in the cybersecurity space and wanted to engage more. But when it comes to the AI angle that was sort of not really a big part of the hearings, the hearings, largely focused on some of the activist investor interests that have focused on oil and gas investing and investing for ammunition manufacturing companies, and whether it’s right to have activist investors play a role in that. And another thing that they wanted to focus on, and sort of, you know, amusing way as well, with Senator Elizabeth Warren calling jamie diamond, the star of the overdraft show was how even though the government is allowed certain amount of relief and certain amount of overdraft allowances to the banks themselves, many of those allowances have not translated into actual benefits for consumers.

JJ Hornblass
So the rain Myra, do you when you hear kind of how the senators dressed some of the AI automation, technology dynamics? I mean, to what degree are they getting to the heart of what we see as as central issues or even challenges in the market? versus maybe skirting around the issue? Meyer, do you wanna maybe start there?

Myra Thomas
You know, I think that officials are notorious for not really understanding automation and technology that, you know, inside of businesses, no matter whether it’s financial institutions, or what have you. So I think much of the challenges that are faced by consumers that they mentioned, yeah, they’re, they’re real situations, they’re real things to think about, not just in cybersecurity, but I think they’re many other challenges that, you know, simply do not come up, because there are certain issues that are, you know, much more attractive, less sick. And so there are other challenges to automation certainly will never come up, you know, whatever it might be.

Loraine Lawson
Lorraine, you have a well, I was particularly interested when Jane Frazier said that city had developed one of the first world’s first companies to develop our own ethical principles for artificial intelligence that really stood out to me. And I would have would like to see more questions about that or know more about that I’ve been reporting on ethical AI. And the bottom line is the experts don’t know exactly how to create ethical AI, it’s very much a developing area. So I’d be interested in learning more about that and seeing some more pointed questions along that line.

Myra Thomas
Absolutely. I’m always curious when someone says, first to establish any sort of automation. So I would love to know it, too. I mean, I know we’re working on an ethical AI piece right now, which I think is pretty much focused on credit underwriting and sort of thing. But, yeah, it would be interesting to know how they can actually document something like that.

JJ Hornblass
Do you even Would any of you even expect them to actually publish that, though, that guidance

Loraine Lawson
would be helpful if they would? You know, you know, because I know I talked to Gartner about this and talked to Queens College and professor at Emory. And they all said, this is a very developing area. And it’s there isn’t even so much best practices as next practices, they called it. So love to hear what they’re doing to ensure that the AI is ethical.

Myra Thomas
I think what’ll happen is like with any sort of regulation that sort of comes down like it’s consistently altered, and then banks will have to play sort of catch up game after they’re installing and going through, you know, all of these really expensive. It rollouts, then they’ll have to find themselves constantly tweaking systems to sort of comply, do what the government wants them to do.

JJ Hornblass
We’ll see you if you think that they’ll eventually be government guidance on ethical AI Myra,

Myra Thomas
I think they will eventually I mean, I don’t know what that will, you know when that will play out. But I eventually see that happening at some point. Because, you know, once we get to a situation where there’s enough banks employing AI in some sort of way, it’ll come up as far as the government regulation. And

Loraine Lawson
it’s interesting, Joe Biden is actually creating an office of AI, sort of a new thing that’s happening. So it’ll be interesting to see if that also plays a role in in developing maybe some rules around it. I’m sure regulators will get into this game. We know, in fact, the OCC and several other organizations of the Federal Reserve included, are looking at what they issued a request for information to financial institutions and the public to find out more about it. So they’re definitely looking in that area.

JJ Hornblass
Let’s turn our attention to at least one bank that has that is deploying RPA in a more significant way. Myra, give us an update on what’s happening at Key Bank, which has about 176 billion of assets.

Myra Thomas
Yeah, I had the pleasure of talking to Dominic cugini, who’s their chief information officer of service, digitization, and Key Bank, and he was pretty forthcoming in how the organization had rolled out RPA. They have been working with automation anywhere, I believe, since 2017. And it was interesting to find out that, you know, the bank is really much rolled out our pa processes across the organization. So we’re talking about commercial credit, every, every sort of part of the organization that you can think of frontline middle back office. And it was interesting, because the the first thing that I talked to him about, because it’s sort of come up many times is what the return on investment was in, in regards to this rollout. And like many banks, they don’t necessarily want to talk about, you know, that return on investment. He was very honest, and did say that, you know, they definitely had a very, very large return on investment, but that it was not initially as much as they thought. In fact, it was a savings, but none at all what, what it should be for the investment. So they went back they retooled, they utilized process mining, to really look back at what needed to be automated and what did not need to be. And were able to improve that return on investment. I think it was interesting, because he was also talking a bit about how they’re going to be utilizing low code and hiring more people on their development teams. In order to facilitate more RPA rollout, as well as looking into workload options, I think they’re looking to add some sort of platform in that. And shortly, he was also very forthcoming and saying that they weren’t actively hiring for those squads. So that was interesting to know. But the one thing that I thought was even more interesting was the fact that, you know, as far as a strategy play as you’re using vendors, picking automation anywhere, and then picking, you know, their best rival in working with UiPath, to basically provide them with the same products. And so they were playing each one off against the other and deciding, you know, basically, what parts of the organization they would use, you know, their automation and what other parts would be used, you know, from the other vendor. So that was interesting to hear. Now, they’re also looking to add other organizations to provide them with OCR tools. So, you know, like many, many banks today, I guess the best ones are realizing that they have to deal with a variety of vendors, of course, with the addition of the variety of vendors, then you add in, you know, significant third party risk, but, you know, at the same time, you know, it’s a business use case, for an organization that keeping

JJ Hornblass
or jazz. Does this remind you of any other particular financial institutions? Do any others come to mind that compare to keep bank?

Myra Thomas
I mean, generally speaking,

Jaspreet Kalra
if we’re talking about the surgery, lowering in return of investment expectations, yeah, that’s been the case with multiple institutions. And that’s what we’ve been hearing from analysts as well, that when the product first rolled out, everyone got really excited. But when they realize that these products also need to be managed and governed, that brings an element of course that sort of cuts in and reduces the amount of savings you know, they can generate.

Myra Thomas
Yeah, I think that’s why process money has gotten very successful, because they’re actually sitting here thinking about, you know, what tasks are we doing and what, whether or not we need to automate them in the first place. You know, he said that they had really, really, you know, sort of shot the moon as far as trying to think about what they were going to actually save. And then when it came back that first year, I think they will that was 2017. You know, the number was way smaller. But he wouldn’t give a number. That was interesting.

Loraine Lawson
No, I actually also spoke with keybank. Coincidentally, Michael Reynolds from their technology department, and he actually was able to put some numbers around their PPP bots, they deployed nine bots for PPP. And of those, they ended up keeping four to use and other loan origination areas. And they said, he said this for bots together, save them, about three and a half or three full time and one part time employee and hours each year. So they did have some successes, although maybe not what they were hoping for. Yeah, I

Myra Thomas
think the big thing with them, he didn’t mention PPP as well. But also, they’ve seen a, you know, extraordinary increase in mortgages, and they’ve not been able to keep up with higher. So they’ve deployed a considerable amount of RPA on that, and, and have been very successful in dealing with, you know, additional mortgages that they’ve been adding in. So that’s been successful for them. So I think that they, you know, gotten buy in from the C suite as far as what automation they’re going to add, and it seems like it’s, you know, quite, quite substantial.

JJ Hornblass
So speaking of value, we had, we shared some tips or best practices for extracting value from bots, rain give us the the five Greatest Hits, best practices for extracting value from bots.

Loraine Lawson
Okay, so the first probably won’t be surprising. If you’ve ever been involved with an IoT project, it’s find a high level executive sponsor. And Annette Kumar, who works with us path is the Vice President of financial services for our P provider UiPath. He said, you know, that he saw in the earnings reports over the last year, that CFOs, and CEOs and CEOs, all the cxos, were mentioning automation, and he said three or four years ago, that wasn’t the case. And we saw that ourselves in some of our reporting and earnings reports. So he said 100% of all the managers that have high use automation have either the CEO or the CFO is a champion. Now a corollary to that is that you should focus on business problems and use business KPIs. That’s where successful companies see the greatest return. So you know, just because you have a budget doesn’t mean you’re now looking for all the places you can use it, but you want to focus on your problem areas. The other and we’ve talked about this before, Myra mentioned earlier, is defining your processes to process discovery process mining, basically, the the act of looking very carefully at what you’re actually doing, and then what you want to be doing. Another sort of surprising, best practice, other I guess, really not a surprising stop Think about it, but is to involve employees in the automation. You know, there’s this sort of urban legend that automation kills jobs, but the more successful companies actually involve their employees in the automation and put the automation or enhance their employees, so they can get offload some of these repetitive tasks that are just time wasters for them. Another best practice is to use bot governance. A lot of people are deploying bots without governing them. So for instance, that creates problems when the bot does it work as it should. So suddenly, you don’t know where the bot stopped when it stopped and you don’t know how you should restart it. Should you started the process over or can you maybe start in 10,000 processes in or you know, transactions in and finally, this is my favorite it’s test tests before you deploy your bots keeper Regina guard who is Associate Professor of information systems and operations management, Emory’s University’s business school, I asked him for one best practice and he gave me that

JJ Hornblass
and who knew that there were urban legends about box out? Lorraine thank you so much that they jaspreet admira also as well. Thank you all for for joining us for this episode of the buzz. We look forward to seeing you next time. Please visit us at Bank automation news.com and on Twitter and LinkedIn. Take care