This week, Bank Automation News shared revelations and surprises from the BA Ignite virtual conference, held Tuesday and Wednesday. The main takeaway: Banks remain in the early stages of automation, pursuing the low-hanging fruit of bot automation, rather than evolving a more comprehensive strategic play, as advocated by PNC’s Enterprise Chief Information Officer Ganesh Krishnan in his fireside chat.
Financial institutions Citi, Synchrony, Stash, OakNorth and Unqork shared how automation is helping achieve efficiency gains, although bots don’t always meet expectations, noted Matt Carbonara, managing director at Citi Ventures, Citibank’s venture capital arm.
The BAN staff also shared their favorite surprises from the conference, including DEMOvation winner Julius Technologies’ shocking statistic that 90% of machine learning models actually fail to make it into production.
Find this and more in today’s episode of the Weekly Wrap, featuring Publisher JJ Hornblass and Associate Editors Jaspreet Kalra and Loraine Lawson.
Bank Automation Risk Summit, on June 15-16, is an all-new event that will explore automation in regulatory compliance and risk management for financial institutions. At the virtual event, attendees will gain valuable intelligence on new practices, methods, processes, tools, and data for improved risk management planning and performance to drive better results and fewer risk failures. Learn more and register at www.bankautomationrisk.com.
The following is a transcript generated by AI technology that has been lightly edited but still contains errors.
Hi, everyone. I’m JJ Hornblass and welcome to The Buzz from Bank Automation News. Welcome. This is where we chart the future of banking automation technology is our weekly wrap for what’s happening in the world on April 16 2021. Thanks first to our advertisers, MX, Narmi and NVIDIA for their support. Thank you to them. And so pleased to be joined by Loraine Lawson, and Jaspreet Kalra of the Bank Automation News team. Hello to both of you. First, let’s start with some general technology news. Yesterday Thursday, April 15. The Dow Jones Industrial Average closed above 34,000 for the first time, reflecting strong earnings and upbeat economic data as a result of the pandemic beginning to wane. The S&P 500 also notched a record close yesterday, this was the 22nd record of the year for that particular index. This week, Amazon Prime announced that it had hit it has hit 200 million subscribers. Amazon also recorded profit profit in 2020 of $21.3 billion. Dell announced that it was spinning off its $52 billion take in VMware and mortgage tech startup better.com b tt er.com, recently landed a $500 million investment from SoftBank, the company’s total funding is now more than 900 million. And this week, we were fortunate to present our 2021 Bank automation ignite conference, hearing from executives from around the FinTech banking and banking automation world, we’re going to spend our time focusing a little bit on some key takeaways from the conference. And I wanted to start with a key with kind of a more of a general feeling that that perhaps you just read in the rain had, what is your general sense for the state of automation technology? In other words, how dedicated are financial institutions to pursuing pervasive automation in their companies? And and secondly, where are they in the continuum? How far have they come? How far do they need to go? Lorraine, you want to maybe kick it off?
Sure. So I thought it was interesting. I don’t think they’re as far along as they want to be. First of all, I think Dinesh Krishnan made some good points about that saying that the or maybe 50 60% as far from being able to automate across processes. So right now where we’re at is what I gathered from the conferences where they’re at is they’re they’re cherry picking and automating some aspects of the process, but not across the process. And, you know, maybe that’s not necessarily where all banks want to be. But, you know, for the bank, like PNC, I think they do want to add their they do want to be able to automate across process all 10 steps as as Mr. Krishna said, So, my fear was, we’re not there yet.
Well, just to just to share, ganache is the enterprise Chief Information Officer for PNC Bank? He was one of the speakers at Ignite. Yeah, it would seem that at least based on what Dinesh mentioned, is that where I guess you know, like the famous Bon Jovi song halfway there were kind of it seems like we’re halfway to really where financial institutions want to go. I think that I, you know, that the discussion around virtual assistants was kind of endemic of how far into institutions have come you had, you had wells, you had Bank of America, and you had TD? No RBC suit excuse me was a word speakers on that panel. Am I saying that right? Megan in that rent? Check? Yeah. So, you know, and so the the notion there was, you know, they’ve made a lot of progress, but have a lot of progress to go. I mean, jaspreet is that, you know, what was your kind of overarching take of where things are at.
So you referenced a Bon Jovi song The way I sort of like think of it as a sports game. So the first half of a sports game is usually what tells you what’s going to happen in the next half, which is exactly where we are with automation stuff right now, because the first innings has been played, when you had the cherry picking aspect of the low hanging fruit aspect of these things that can be quickly turned around that are visible. Now, the second half of the game, I think, really will depend on end to end processes, what scales and what does not scale, and how much extra value can you deliver if you personalize automations. So that’s the conversation I think we’ll see playing out over the next couple of years as sort of a need and desire balancing takes place, the need arises out of the fact that you have all these legacy institutions fighting against newer entrants who are cloud native by default. So they’re out there automated by default in so many more ways than these older institutions have been. And the desire aspect of it comes from the fact that there is increased efficiency, there is some sort of software transparency, that’s also starting to seep in now where it’s easier to audit things in terms of how a machine made its decision. So those two aspects I think, will really come to the fore in the second innings or the second half, or whatever the relevant sports metaphor might be. So So that’s my sort of take on it. First half of it is done in the second hour will be more interesting to see how it plays out.JJ Hornblass
The deal goes straight to the cloud advantage for some kind of, I think that that did strike me as well as as kind of a demark. Right? Either you can or you can’t. Um, I think the question that, that it raised for me was, to what degree can financial institutions pivot or adjust so that they can they can adopt this kind of approach, and have a similar read, because it because this advantage that challenger banks and fintechs have as being cloud native, you know, it really could widen a gap in terms of services and capabilities, and certainly from a cost basis, and, and innovation basis, you know, an ongoing, kind of, you know, iteration level of iteration. I don’t know if they’re, if that’s possible.Loraine Lawson
I think, I think what Mr. Krishna said about, you know, in 2013, they started laying the groundwork with the data lake. I saw this with data, too, when I was covering data management, data integration and data, just everything data. There were companies that thought about it strategically and invested then, and their investments didn’t immediately pay off. For those low hanging fruits, it was a long term play. And those are the companies you’re going to see winning this game to, it’s the companies that are thinking about it strategically, that are investing in the technology that will support it now. And the ones that are just deploying, but I think, you know, they’re gonna have, they’re gonna have a hard road to hoe to catch up later.JJ Hornblass
I guess the issue is, if if there’s a centrality, a central advantage in being cloud native, then that kind of investment that you’re talking about is by default, diminishing the efficiency or results. You know, now you’re gonna say, well, it’s a near term thing, right? It’s a relative, relatively near term, right. You’re making the investment now, to enhance efficiency later, you know, but when you’re a publicly traded bank, and you’re making that investment now that has that has negative implications. From a, from an investor standpoint, that have that could have, you know, could have reverberations later on. And then is that not a fair way to think about it?Loraine Lawson
I mean, PNC seems to have done it, and they managed to do it. And I think, you know, I do think the bigger banks have an advantage here. Obviously, they have, you know, he has 1000 developers globally. But that’s so looking to the cloud is going to be how smaller banks keep up that’s gonna be their only play, but Again, I do think they have to think about it as a strategy instead of a tactical play.JJ Hornblass
Yeah. I always say, Oh, sorry. And desperate. Did you want to say, sir?Jaspreet Kalra
No, just a small thing that I was just speaking to a source last night who’s sort of observation struck me as very relevant when we think about organizations with limited resources. And he said that, when you have limited resources, you really have to pick your battles. So you cannot fight as widely as CEO JP Morgan, or Wells Fargo does, because they can afford for some projects to fail and some to work. Whereas smaller institutions, they want a higher success rate, because there’s less room for failure. So I think that’s why you’re going to see the processes, they will of course, be some leaders in the pack that everyone emulates. But at the same time, I think you’re gonna see this sort of nimbleness also play out with smaller institutions, when they figured out that something was working, they can turn it around much more quickly than a multibillion dollar multi globe. Sorry, multi country organization.
Well, you you even Ganesh from PNC said that they’re not in the same league as JPMorgan Chase, right? When we, when it you know, jamie diamond jpms, CEO this week said that they have 150 ai projects, at live AI projects. And at PNC, that’s, you know, that’s basically 3x what PNC has at present, and, and, you know, at a multiple of that, you know, of other financial institutions as well. So, you’re picking your battles in jasprit, even on a relative basis, to a large financial, you know, for large and even at the relative size of, so this is, these decisions become very profound, which is why I think coming back to something that you said Lorraine, where ganache was kind of, you know, put a number on the the volume of processes that are, you know, have are still paper based, you know, this, which is why I think that, you know, one of my takeaways from the conference is actually document automation automation is probably has more import and legs, then we may be realized after, you know, that we kind of don’t really, you know, it’s it’s not, it’s not sexy, it’s not, you know, kind of the, you know, really cool, um, but it is the document automate automation may have more profound implications for banking, if I’m talking about, you know, hardcore, you know, systemic banking efficiency, then, you know, bots, virtual assistants, whatever you want to call it, you know, I don’t know if that was, if you had a similar takeaway.
I did notice that Mr. Christian talked about that, about their document process, using it with PPP, the payroll protection plan to extract data and information from these millions of documents that they got. And that was something that they didn’t have to hire people to do. That’s not the first time I’ve heard of someone using AI based tools to extract information from documents. Scotiabank, last week we talked about, has an AI tool that has saved them millions, and they use it in the legal department to check contracts, and flag, you know, flagged things that need to be flagged, but it can also correct things that need to be corrected. So I think you’re right.
Yeah. What are what are a couple of, you know, kind of glaring shortcomings if you if you spotted any, in the strategic approaches, that were kind of on display at at our conference this week, I guess other than the strategic investment notion that you shared Lorraine. I mean, one, for example, that comes to mind for myself was the point that Matt carbonara from Sydney ventures mentioned, which is that exceptions in processes really remains a sticking point and, and really a sore topic. It sounds like from Citi group, where, you know, those exceptions ended up you know, costing them I don’t mean in dollars, but in in, in process and effort. Great, you know, much great much more than they expected when they installed automation. What were there any kind of glaring, you know, shortcomings or or needs that that were expressed during the conference?
I mean, I’d say so as you said, the exceptions thing is something that still stands out quite a bit. And the story I always go back to on that is if you use a phone to scan a room, if it’s just regular furniture, it will be just fine doing that will understand what’s what. But if you put a toy elephant in between, it gets completely confused, because it doesn’t know how to account for that variable has never encountered. And on the other end, I think what sort of stands out is this demand for end to end solutions, wherein you just don’t have a small part of the process being automated or a small bag of it being automated, and yet, you have human surrounding that process. And the other thing was this need for domain specific knowledge that our computers labeled or software is able to bring to the table instead of being a general application thing. And that gets tricky, because as I think Matt only made this point that it is much easier to teach a computer How To Play Chess, then how to, you know, navigate the open world, because there’s a lot more unexpected variables in the opener. So those are the sort of challenges that will come up when AI and ml come much more forward in these technologies. But I think there’s always also ways to solve for that which are being experimented with
Yeah, I heard two things that stood out to me. One is the need to automate when you automate to automate risk compliance and auditing. And I’m not sure that’s being done. The other thing that I heard that was I thought was very interesting, from green.is, the way that they have a lot of the conference focus on personalization, and how you can use these tools to personalize and they have personalized their products, you know, used to personalize product approach that I’m not, you know, hearing from banks as much like their personalized recommendation chatbots that do financial recommendations for people or to sell them products, but that’s different than personalizing the product itself. Right. So I thought that was a fascinating thing that maybe, you know, there’s an opportunity for some innovation there.
Yeah, I think this, this brings me to my last question, which is, one, what is the most surprising thing that you learned at the conference? And for? I’ll kick it off. I thought that the most, that the most surprising thing was the degree to which virtual assistant data is going to end up being such an advantage for us for financial institutions, that that that mine of data, it’s kind of to your point, Lorraine, right. And which is that, you know, that that mine of data is is going to end up propagating products, a customer experience, overarching strategy, resource allocation for financial institutions for, in my view, many years to come. This was to me, kind of shocking revelation when you think of all the data, for example, that Bank of America has extracted from its Erica virtual assistant to date and is sort of still amassing by the nanosecond.
Yeah, it’s similar to that. I thought it was interesting when MSU Federal Credit Union talked about its filing a financial conceir that maybe we’ll even not only be able to use that data, but we’ll be able to ask consumers for more information. So that was a powerful of realization that that sooner chatbots may be quizzing us about things but for me, the biggest thing that I shocker to me was actually during our dinner demotivation. The winner Julia’s technologies, their CEO said that nine out of 10 machine learning models fail to make it into production. Because of data problems, I found that that to be incredible. And incredible statistics so I was interested in that.
That is that is interesting learning. Just be what shocked you.
I mean, I wouldn’t say shocked me, but it was
something must have shocked you.
But so so I would say that the sort of standout element of that for me is that when we think about new technology, a lot of is sometimes end up getting tunnel vision in the sense that If this is the future, this is how it’s gonna roll out. But at the same time, you come back to banking as a relationship, this predates business. And you, that’s where a lot of value is generated in this business. That’s why I think that this entire idea that was made by the representative from keys bank, Justin Hansika, that you’re gonna have this combination of physical and digital experience, the things that you can do at home, please do at home in front of a computer. But when you want to come talk to someone, get some real advice about things that matter. We’re here for you. So I think that sort of transition from Okay, you go to a bank, and you wake up on a Sunday morning or Monday morning, like oh, my God, I have to go to a bank, that becomes a very different experience in the sense that you have to know exactly who you’re going to meet their you know, what you’re going to talk about. So it’s a much more seamless experience. So I think that’s going to be the surprising change that happens slowly within banking. Like it will be an undercurrent, but it will also transform the way the industry sort of thinks about itself.
Very nice. Very well, thank you both for your insights, really great talking to you about it. If you if you miss bank automation ignite, there’s plenty of there plenty of stories from the conference, there are recordings of many of the sessions. There were some excellent sessions at the conference, and you should certainly check them out at Bank automation news.com. What else should folk expect from us over the next week or so?
So I’ll be looking into stories. Some of them speak to some of the trends we talked about how robo advisors are becoming much bigger and how they will supplement or supplant. That’s the hypothetical in how will they supplement or supplant human beings within the banking business. And then we’re also working on sort of like a in depth forum or features should look on how personalization is affecting automation technologies going forward.
Great. So again, please visit us at Bank automation news calm and follow us on Twitter and LinkedIn. Were so appreciative of you joining us for this episode of the buzz and we will see you next time. Take care.