• • Scott Blandford, Executive Vice President, Chief Digital Officer, TIAA

Download Podcast
Apple Podcast, Google Podcast, Spotify, Pandora, iHeartRadio, SoundCloud, TuneIn, and Stitcher. Find other syndication channels here or search CIO Talk Network podcast on any other app.
Explore More

• • How AI is Transforming the Financial Advice Business
  • I & O Investments: Technology or Operational excellence?
  • Should Business Influence Or Control IT Investments?
  • Residual Value of Business Continuity Investments
  • Steps to Enabling a Frictionless and Future-Fit Finance Function
  • Adopting AI and Machine Learning in Healthcare
  • Trusting Artificial Intelligence for Big Business Decisions

Transcript 

Sanjog Aul [00:00:22]:
Hello and welcome to CIO Talk Network. To learn more, please visit ciotalknetwork.com and the topic for today is ‘Should You Continue to Invest in AI and ML?’ So with COVID 19, which is a healthcare crisis, it went down, but it’s coming back up again and it’s also spawned a financial crisis. So there are some troubles that we are looking at. Yes, there is always a silver lining in the cloud, but still we have to deal with issues. So with that said, just because business has to go on, there is slashing of the costs. People are hiding, but to a very limited degree, not doing projects the way they want to, and they are even halting hiring of any new talent. Then you have some other issues which are keeping people up at night, how to deal with business continuity, etc. With all of that said, let’s talk a little bit about the AI and ML initiatives, which gained a lot of investment and momentum and interest among many communities, including the investors and the IT leaders and the business leaders, so that they can get the most of it.

Sanjog Aul [00:01:33]:
Now, given where we are today, given the uncertainty, how should organizations be looking at, rationalizing and understanding, what would it get if we were to invest properly and prudently with AI and ML moving forward? So to discuss all this, we have Scott Blandford, Executive Vice President, Chief Digital Officer with TIAA. Hey, Scott, how are you?

Scott Blandford [00:01:59]:
Hey, doing great, Sanjog, how are you?

Sanjog Aul [00:02:01]:
Very good, sir. Pleasure to have you. So, as I laid out the interesting times we are living in, what do you think?

Scott Blandford [00:02:09]:
Yeah, no doubt.

Sanjog Aul [00:02:10]:
Yeah. So with that, I’d love to at least see, if historically we started on this AI and ML journey a couple of years ago?

Scott Blandford [00:02:18]:
Right.

Sanjog Aul [00:02:19]:
It’s new, but still a lot was done. What do you think has happened so far with that investment? Are we still in the sandbox or we saw some organizations getting some tangible value? Where do we stand today in your view?

Scott Blandford [00:02:34]:
Well, you know, I can speak on behalf of what we’ve seen, which is, we’re very happy with the investments we’ve made and we’re going to continue to invest pretty heavily in this area, and part of that is driven by how we’ve chosen to deploy. So the way we look at it is, job number one is helping customers be successful in whatever they’re trying to do, and we’re a firm at TIAA that is focused on helping people get through retirement. What that led us to do with AI and ML is focus on very practical use cases that help real people with real things. So we’ve leveraged it heavily in trying to understand through our automated assistant what customers are trying to do. So we offer full range of services on the web, on mobile. Our AI ML journey has been very focused on conversational UI and as somebody is typing in a chat screen, not just getting answers to questions, but performing, you know, financial transactions on their account.

Scott Blandford [00:03:45]:
It’s quite complicated to understand from a few words and numbers what people are trying to to do. So, you know, we started in the sandbox, but pretty rapidly deployed out to real customers. Our automated assistant, which has had fantastic uptake, fantastic growth. So we’re going to continue to invest and expand the range of what our assistant automated do.

Sanjog Aul [00:04:11]:
So, you as an organization are going to try to work towards making investments. Even the era before the AI and ML, some incremental tweaks were being done. So this did give us some promise, right? AI and ML was supposed to revolutionize and some people thought about that as, a people displacement risk, if you will. But then slowly people started recognizing that it might actually help people do better and bigger things together. Now with all that said, would you say we have enough fodder to take it up to the management that come what may, come covid, come any crisis, we cannot stop, we should not stop investing?

Scott Blandford [00:04:59]:
Yeah, I think so. Our firm, we’re very oriented around, taking care of customers is job number one, and we know that our customers are looking for different things, especially in the time we live in now. One thing they’re looking for is just help with understanding their financial situation. They need a partner they can trust to help them make good financial decision, and we’re a firm that can provide that. So, we want to make sure we’re there for our customers. The other thing that’s happening is everybody’s remote and people are home.

Scott Blandford [00:05:39]:
So they’re looking for increased access to digital services and consequently they’re looking for natural language, and our services are supported by AI and ML. So we’ve seen prior to the current pandemic and everybody being from home, we were seeing pretty big uptake anyway. But this has just taken it to a whole new level. So we’re getting great support in our company for continuing to invest here.

Sanjog Aul [00:06:10]:
So would you say AI ML turned out to be a better mousetrap or something we never thought possible and it got enough rave reviews that no matter what, you will still be able to show an ROI formula which stands the test of time, including the times that we are living in?

Scott Blandford [00:06:29]:
Yeah, I think so. This may be an overly simplistic way to look at it, but one way to think about it is a digital customer. So a customer that’s comfortable working with their technology. Some people prefer web, some people prefer mobile, some people prefer Alexa, some people prefer Siri, some people prefer chat windows. What we’re trying to do is make sure wherever our customers are, the way I try to think about it is we live in the attention economy where everybody is starved for spare attention because everywhere you go there’s a screen or something reaching out to you, saying, please pay attention to me. With all the great content and exciting things people are directing their attention to. If we’re lucky enough to have somebody spending a few seconds or minutes thinking about their financial well being or their retirement, we want to be there in that moment. That’s the best way to help people is, when they’re thinking about it, we want to be there.

Scott Blandford [00:07:38]:
If they’re thinking about it near a chat window or on our voice response system, on the phone or Alexa or Siri or the web or mobile, we want to be there. So I think, yeah, for sure. It’s opened up our digital services to a whole bunch of customers that had not been as interested in our other interfaces.

Sanjog Aul [00:08:01]:
Now I use the word ROI. Do you think we should start looking at the ROI for every investment? Of course we are making investments in all different areas. But taking the context of this AI ML, what did you think when your organization, not you, means your group think when they came up with the investment strategy and some formula to say this is what we’re going to get, some hard ROI, some soft ROI, and now given where we are, maybe you would tweak the direction or the magnitude to which you will invest, but you should have some form of ROI calculation which you would perhaps calibrate given the times we are in? Are you first of all, and if yes, then how?

Scott Blandford [00:08:45]:
Yeah, I would say when we embarked upon the journey here, we started with this from the standpoint of, we want to deliver real practical help to real people, and if we do that, it should have a really great ROI. So we put our business case together and made the investment and we’re handily beating the case. I will say that at TIAA, we look at this over the long term and our company has been around helping customers for 100 years. So we try to think of what’s in long term best interest to customers. What’s the long term best interest of our firm. So our business cases typically, can have extended payback periods if we believe in the thesis. But I will tell you in this case, it hasn’t taken a long time.

Scott Blandford [00:09:43]:
Like the team that has deployed our AI ML, have done a fantastic job and customers are getting terrific benefit from it every day.

Sanjog Aul [00:09:54]:
You seem to be a pretty good evangelist for AI ML, so that’s good in your role. So yeah, that’s your job to go ahead and evangelize but how well is your management biting? Or I would say, you talk about your organization, but what do you see the appetite for management, which is not directly going geeking out on AI and ML per se, or are responsible for the technology functions or digital functions? How well are they able to digest something which was never understood or was not even existing earlier and now it brings capabilities they never imagined earlier and suddenly there is a stop and there’s a reason why they should not invest there, and instead of focusing on AI and ML, they should focus on keeping the lights on? What level of enthusiasm do you see among those members of the management?

Scott Blandford [00:10:48]:
Yeah, Sanjog, I think we look at it like this. It’s not about the technology for technology’s sake. It’s about helping people. This is technology we use to help people in a new, different and better way. So there’s tremendous appetite in our firm for that and they understand that there’s newer technology attached to it. But we sort of stayed away from doing tech for tech’s sake and have very much focused on let’s help people. The times we’re living in now, they need more help than ever.

Sanjog Aul [00:11:28]:
So it’s been not a battle, not a convincing exercise. Is there something that you did or your team did, including maybe working with your CIO and others to explain the value and did you show some value upfront so that now the belief system is set in?

Scott Blandford [00:11:46]:
Yeah, I would say, we did two things that have led to that. The first one is that for sure, our business leaders in North Harm champion technology that helps customers, our plant participants and our plant sponsors. We work hard to bring everybody along on the journey. So just bringing people with us, making sure we explain what we’re doing, that’s like the first thing, and the thing we did is, we’re very deliberate about the risk envelope for things we undertake. So we started very small. Let’s see if we can deliver something here for very narrow range of customer needs. Let’s get that working really good and learn a lot before we scale it up.

Scott Blandford [00:12:38]:
That way the initial investment, if you go back a couple years, was very small. What we’re trying to do is get smart enough to see how we could scale it up. So, low and behold, it’s a few years later we’re scaling it up and every dollar we invest goes a lot farther now because of the work we did initially. So I’d say those two things, bringing our stakeholders along so they can be champions for us, which they’ve been, and then starting small so we could get smarter ourselves so that when we’re ready to scale it up, the ROI cave is really good.

Sanjog Aul [00:13:16]:
Would you say your organization in particular or your industry, did it get hit to the degree many other industries did with covid?

Scott Blandford [00:13:27]:
Well, no. We felt it like everybody else, we felt it with the people that work at TIAA, moving to remote work. But we have built such a strong foundation and the last however many years here that when Covid hit, one of the things the government did is they changed the rules around retirement plan. There’s a piece of legislation called the CARES Act that allows you to tap into your retirement plan savings for urgent needs. Full support of the CARES Act is I could take additional kinds of withdrawals and loans from my retirement plan. I can suspend repayments from prior loans, etc. To allow people in tough time access or saving temporarily. We are proud to have been the first retirement provider on the street to launch full digital support of the CARES Act basically over the weekend that it was introduced and approved by the government. So look, that’s not an accident because we had done the foundation work and we’ve got a really solid architecture, solid partnerships are around the company and our firms really driven by helping customers.

Scott Blandford [00:14:58]:
We knew that in this time of uncertainty, the government introduced the CARES Act. It’s passed. We want to be there Monday morning with full Support for what folks are going to want to do. And we’ve seen terrific uptick of the services and that includes standard digital services, web, etc, but also with our AI ML.

Sanjog Aul [00:15:24]:
Now think about the companies. Of course, I’d say you’re fortunate that you did not see that big a hit and that’s why you are able to maintain the continuity. Talk about industries, hospitality and others who also could very well use this newer technology, the computing paradigms, to create value. But their business situation is so dire in many cases that’d be tough for them to do. What would be your advice to such leaders working in such industries?

Scott Blandford [00:15:59]:
Well, I’m not sure I’m the best position to give advice to the other industries. I just know that what’s worked for us is staying focused on being practical and staying focused on helping customers. I don’t know what that means in other industries, but I know what it’s meant for us and we’ve really maintained our focus there.

Sanjog Aul [00:16:27]:
So you being the Digital Officer, is there a vision against which these investments were made? Is there something that you’re going after and was that at all disrupted with covid?

Scott Blandford [00:16:41]:
Well thanks for that question, Sanjog. The way I think about what we’re doing with digital for our firm is really five things. Number one is we’re trying to make sure our customers have great service experience with whatever product feature they’re accessing. So if they’re looking to pay bill, use online bill pay, if they’re looking to make investment changes, if they’re looking to retire that, every service and feature we have, it’s got to be straightforward and easy to use. So we have been doing and have continued to work there. Second thing is at our firm we provide a full range of financial services and we strive to have, one plus one equals three here, so that there’s value to customers of having multiple financial products and in one spot, easy to move money from your checking account to make an IRA contribution, it’s easy to make your mortgage payment from your checking account, etc.

Scott Blandford [00:17:54]:
So we think there’s a lot of value in the fully integrated experience, and we’re one of the few financial firms, I believe out there that has the full range of products, all integrated in one mobile app and in one website. There’s a couple of products that aren’t in the web today, but they’ll be here in the next few months, it’s all together in one spot. It’s not multiple websites, not multiple mobile apps. That number two is fully integrated experience. The third thing we’ve been driving forward on is, we know financial services, complicated, sometimes opaque, and customers need help and sometimes they need help with a quick little bit of content to explain things to them. Sometimes they need help with a little wizard to guide them through a decision making process. Sometimes they need to talk to somebody on the phone, sometimes they need to do a video chat.

Scott Blandford [00:18:49]:
We call that advice everywhere. We want to make sure we’ve got advice, some kind of advice everywhere a customer might need some help. The fourth thing is, and this is very germane to our conversation here, is we know customers are increasingly in disparate places. They’re on their phone, they’re using a voice Alexa or Siri Assistant. They’re out there in the world on different devices. So we’ve got to go be where customers are. So it’s like, when I talk about this internally, the old joke about, what’s the most important thing in real estate? It’s location, location, location. Well, it’s the same thing. What’s important in digital.

Scott Blandford [00:19:37]:
We gotta be wherever our customers are. Then the last thing that’s part of our strategy is with all the power and features that providers like ourselves have deployed on their digital products, web, mobile, you name it, there’s just so much there. It used to be on the early days of digital, more was more and customers were hungry for services so they were eager to use whatever you would maybe deploy. What we see now in the attention economy is, it can be overwhelming for people. So we’ve put a concerted effort on what I call digital lightness. Digital lightness is about stripping away unnecessary clutter in the experience and that we don’t want to offer people things that there’s some chance they might use. What we want to put more front and center are things that it’s highly likely they will use.

Scott Blandford [00:20:40]:
The full range of features and services we offer to everyone, but the emphasis is really only on things that we believe are highly likely a customer will want to access, and then everything else gets pushed into the background. That’s digital lightness. So back to your question, hey, is this fitting in an overall framework? Absolutely. AI ML is part of advice, helping people make good financial decisions. It’s part of emerging channel and our outreach being where customers are. That’s the conversational user interface.

Scott Blandford [00:21:19]:
It’s part of digital lightness, making the decisions about what content we should emphasize and what content we should suppress so that people don’t feel as overwhelmed when they’re working with us. So it’s very important to the whole strategy and those just a few examples. Hope that makes sense.

Sanjog Aul [00:21:37]:
Absolutely. Let’s take a quick break listeners. When we are back, let’s look at the environment that we are in and TIAA is fortunate that they have not been hit as hard. But then there are testing times regardless and even though as a business there may not be much of a hit, but then people are still working remotely and there are some other issues across the value chain partners and even the customers that we are dealing with. Can AI and ML come to rescue? Can it bail us out during these testing times? Let’s explore when we come back. Please stay tuned.

Sanjog Aul [00:23:03]:
Welcome back. So Scott, there are a lot of different companies and in your role as Chief Digital Officer you got your counterparts in other companies and the CIOs in all different business leaders and not everyone is in the same environment or same situation. If you were to not grossly generalize but still share creative ways where AI and ML could help bail them out during testing times or justify its investment in itself, what would those areas be?

Scott Blandford [00:23:42]:
Well, I would say that the first thing is just recognizing it’s a very challenging environment and no one really knows how long the circumstances we’re in today are going to last. So I guess the number one requirement is being nimble. I know we’ve put a lot of energy in the last few years around Agile and Scaled Agile Framework. If you’re familiar with Agile Manifesto, item number four in the Agile Manifesto is, we value responding to change over sticking to a plan. I think those are the agile skill or what we’re all called to employ now. So one thought is to make sure you’re supporting agile and you’re in your area, whether you’re way down the agile journey and you just need to double down or your team’s at the beginning or partway through it, because the agile where it really shines is in times of uncertainty, responding to change versus following a plan. I also think it’s real important to know where your value is as a firm and making sure you’re staying focused on that.

Scott Blandford [00:25:06]:
In the new economy we live in, it’s very easy to work with partners for the things that aren’t differentiated value for your firm. So you taking an accounting of what are the things I’m doing today that I can work with a partner to do that would allow me to focus my always scarce and precious resource on the things that are special about what my firm does and what my customers value. Maybe the last thing is staying focused on customers because ultimately they’re the reason we’re all here, whatever business you’re in. They need help in different ways today than they did before, but they need help in some of the same ways as before. So just stay in focus there. So maybe those three things, being nimble, doubling down on agile, knowing where your value is and being pretty sober about what you do in house versus what you work partners on and then customer’s the reason we’re all here. So let’s make sure we’re taking care of them, all day, every day.

Sanjog Aul [00:26:14]:
Yeah. As I introduced the discussion today, I did mention about, silver lining in the cloud. Do you think while there is a big cloud out there with covid but could AI and ML serve, or could this environment result in some good, interesting use cases which may reveal themselves as use cases which would allow us to use the potential of AI and ML to a totally different degree or a level? Anything that caught your attention?

Scott Blandford [00:26:50]:
Yeah, Sanjog, I think what we’ve seen is a big uptake and customers accessing us digitally, as you would expect and I think, we’ve seen that across financial services for sure. I saw a stat the other day that 40% of digital banking customers are new to digital banking since the start of covid. So you know, customers are changing with this change and they need more and more access or they’re looking for more and more access to digital services. The folks that are newer adopters of digital don’t have the experience of the last 10 years of digital with web and then mobile. So they’re coming to it for the first time. That makes sense, right? They’re all new digital users. So what we’ve seen is. I don’t know that this is a silver lining, but we’ve seen a lot more access for these newer services that we created than we had in the past. A big uptick there.

Scott Blandford [00:27:58]:
So we’re happy that we’re able to serve more customers in more ways at a time they need it most.

Sanjog Aul [00:28:07]:
Seem, even when times were okay or normal, people were struggling in making use of these newer technologies because the fundamental readiness that you need in an organization with respect to data and understanding of what is AI and how to use ML, some people were deer in headlights, other people were farther ahead and some people were in between. That hasn’t changed. So would you recommend staying away from AI and ML, especially now for the people who are kind of deer in headlights, or is there some hope for them?

Scott Blandford [00:28:38]:
Plenty of hope. The way I look at it is all depends on having good use cases. If you can find something in your business or your area that’s got real practical value, that’s always what you want to do in technology, is build things that are valuable for people. AI and ML are a super powerful tool to make that happen.

Sanjog Aul [00:29:08]:
Now, given all the different departments that we have, you talk about collaboration or HR or supply chain finance at a business level, they are being rethought. So would you say if at some point we thought of AI and ML becoming the very DNA of the organization, would it help? Or would we have to rip apart the original DNA definition of how we will use AI and ML because even the business is being rethought?

Scott Blandford [00:29:44]:
You know what? I know there’s some cases here in our firm where we’ve taken some tasks, maybe observation tasks, where you’re combing through vast amounts of data looking for something. We’ve used AI ML to help make that much easier. So it basically does a lot of basic observation here and then surfaces likely candidates to a real person. What that’s done for us is it’s allowed some of our people to really become like superpowers, right, where they used to be able to do X number of these things in an hour a day, and now they can do 10 hold that because they’re partially powered by ML. So I do think that’s allowing people to begin to rethink the work. We’ve had similar success with robotics. So yeah, I think so.

Scott Blandford [00:30:48]:
It hasn’t caused us to have to rip apart anything. But just as you find new ways to scale processes, you find new opportunities and all of a sudden your people can be focused on things that are higher up the value chain and doing things that only people can do.

Sanjog Aul [00:31:10]:
So what definitely you answered is if AI was or ML were to be used at the very foundational level, it would not have researched things. But what we are referring to here or what we are dealing with is an environment which is fundamentally forcing us to rethink HR, how we will deal with people, how we will have supply chain, and that I’m referring to Covid or any other related ripple effect in terms of the crisis we are dealing with now. When the business shifts, you don’t let the tail wag the dog. So you don’t want aim to redefine but instead of businesses saying this is where I’m getting reset, are we going to do anything different with the AI ML?

Scott Blandford [00:31:53]:
Yeah, I think that’s part of having a strong foundation, getting in these new technologies early enough with small projects so that when the opportunity presents itself, you’re ready, because of the work that we’ve done over the last few years, using our automated epiphany as an example to be ready with it, we were able to be there when the CARES Act launched, for example. So I think it it’s smart to start small, de-risk the tech, learn it enough so that no matter what happens in the business, you’re able to take advantage of the opportunity.

Sanjog Aul [00:32:42]:
Now even that you just mentioned about. So when you created that, but then there was also an impact on the people. Remember I mentioned the displacement part of it, the perceived fear. One is people are anyway concerned because businesses are not doing as well and there are layoffs like there is no tomorrow. On top of it, we bring AI and in some companies they are not fully there yet where people have let go of that fear that AI is going to come and displace us. So that could become a double whammy for such organizations. How does someone deal with that and instead turn AI and ML related proposition on its head to say now this is the time which will actually help save your job versus displacing you?

Sanjog Aul [00:33:30]:
Are there any avenues there?

Scott Blandford [00:33:32]:
Yeah, the way we’ve been thinking about it and doing it is just continuing to upskill our teams, so that as we find new ways to automate stuff lower in the value chain, that our people can be deployed solving problems higher up in the value chain and we’ve seen great success with that. We’re running a big program now helping our technology team move through the journey to get everything in containers and to be able to operate and fully automated releases. The next gen of automated DevOps. Those examples is not necessarily AI ML, where, I love this Marc Andreessen quote, ‘Software is eating the world’ and as software finds its way to automate things, you want to have a company and teams that are able to say, okay, great, that work is now automated. How do I move myself and the work I do further up the value chain and do more for my customers? So we’re trying to bring people along for the journey here.

Sanjog Aul [00:34:49]:
Now with the people, this is your perspective on how you could bring people along. How are people in your organization or you may be talking to your counterparts in other companies. How are people dealing with this crisis and then as you present opportunities for them to “enhance their capabilities” through AI and ML, are they buying it?

Scott Blandford [00:35:14]:
I think so. We’ve got a company that very much cares for the people that work here and our customers. So we’re a mission driven, 100 year old mission driven firm that’s here to help people who help others. That permeates pretty deep in the culture. So one of the things we talk about is making sure in the challenging times we’re in, that we all stay connected and we’re looking out for each other and helping support each other to keep our family and friends and colleagues, safe and healthy. Then where we have opportunity to do some of that upskilling now if there’s an initiative or something that got put on hold, well, now’s the time to invest in making sure you do some learning on AI ML or we’ve got some internal courses we just created on how we do containers at TIAA. So now’s a good time if some of the business stuff is slowed down, to invest in yourself.

Scott Blandford [00:36:22]:
Because we know software is going to continue to permeate every aspect of business and we’ve got to make sure our folks are brought along. So I think people feel supported on that journey. We’ve seen them willing to take advantage of the opportunity and we’ve seen everyone looking out for each other.

Sanjog Aul [00:36:43]:
No one is of course in your organization. Not sure how many other external ecosystem partners you may have. You may have a few in some other industries, there are a lot more. If an organization goes on this AI & ML journey, would you say it is primarily only looking inwards and just fixing or cleaning up the house inwards, or should it permeate across the value chain for it to be truly effective?

Sanjog Aul [00:37:12]:
So if it has to permeate across the value chain, but those people are not, or that those companies we don’t control, how do we bring homogeneity in this?

Scott Blandford [00:37:26]:
Well, I think there’s opportunity across the value chain. The most straightforward place to start is inside your own company and your own division to learn, get comfortable with it and see it working. It’s all theory until you actually feel working live for customers. Absolutely, there’s opportunity across the value chain with partners. Like you mentioned, every partner we have is at a different stage of maturity with it and I think what that does is, as you’re deciding who your partners of the future are going to be, the ones that have made the right practical investments and are thinking ahead, they’re going to be providing, leading services and better value for every dollar we spend with partner, and they’re going to be the partners for the future.

Scott Blandford [00:38:17]:
So I think that’s a problem that works itself out over time, and you only know where a partner is when you go talk to them. We’ve done a little bit of this. I would say there’s more to do here, but a dialogue with a partner around, hey, what new things can we do together? Because this technology on both sides here allow us to do it.

Sanjog Aul [00:38:42]:
Let’s talk about security and in your business, I’m sure security is top of mind for everyone, and when you try to handle things where from a security perspective and that a human is doing certain thing and humans are also not the most logical at times, so they bring their own flavor off security risks. But with AI and ML, you could make one mistake or there could be one hole created and it could exponentially increase the possibility of damage. COVID 19 also disrupted the way we work, our workflows and many other things and drilled holes into our forts. By introducing AI and ML, are we compounding the problem or are we fixing it from a security standpoint and compliance and other risk management standpoint?

Scott Blandford [00:39:41]:
Yeah, I think to your point, AI ML allow bad guys to scale up, but they also allow us to scale up and we’ve seen a lot. Obviously I can’t talk about details here, but a lot of value in us being able to cover more complex use cases, more scenarios, more places to make sure that we’re staying ahead of all that. So it’s been very valuable for us there.

Sanjog Aul [00:40:14]:
Who should own these initiatives? Who was owning earlier? And you’ve said, let’s take your companies as an example, when AI ML was brought in, maybe it came through the IT route and then you got involved being the Chief Digital person and of course you had a set of people who were jointly owning it, what’s that structure which will keep us sane because there are endless opportunities if you want to go invest, and now that you got a disruption in the way business is to be looked at, who should own it now and moving forward?

Scott Blandford [00:40:51]:
Well, the way we try to do things like this is especially for newer technologies, multiple areas in the company can take a look and do some early learning. But then as it matures and we want to deploy to customers, we like the structure where there’s a central team that owns a “platform”, but it gets deployed out to customers and users through various business channels. So there’s central ownership of the platform but distributed ownership of the application. That’s worked well from us, both from expense management side, but also pooling our resources. So group A has an application need, we put that in a common platform, then group B get the benefit of it. So, everybody benefits from the learning each team has.

Sanjog Aul [00:41:50]:
There is enough leeway for all parties that if we don’t hit the mark in terms of the ROI, you have the immunity, if you will, to keep charging ahead and this is still a sandbox for you?

Scott Blandford [00:42:06]:
Yeah, I think that it’s really more that we try to be very practical and back to what I was saying a little earlier, Sanjog, and start small so that there’s not a lot of risk of missing the ROI because we’ve taken the prudent steps up front to be pretty confident.

Sanjog Aul [00:42:31]:
What education did you have to impart? You means, whosoever started this journey for you to be able to get them ready, if you will, to embrace something like, AI and ML?

Scott Blandford [00:42:42]:
Yeah, so there’s general education on the domain and then we work with partners for some of the tech here. So then there’s specific education that the partners brought to bear here and I will mention here but we’ve been beneficial. We’ve had the benefit of working with some really good partners in the space, both with the platforms they brought, as well as helping to bring our teams along, and that’s great.

Scott Blandford [00:43:10]:
Everybody wins. We’ve got very successful deployment. We’ve helped partners with their business, and my teams are smarter and they’ve got real material accomplishments they can point at. So it’s been a great, fun journey.

Sanjog Aul [00:43:28]:
I’m sure it was a great journey, but I’m sure there are some interesting times, right? That’s the word I would use.

Scott Blandford [00:43:35]:
Yeah. Oh, for sure.

Sanjog Aul [00:43:36]:
So could you share some of those interesting instances without taking names or without giving away too much on what are the things that people have to be careful of, the gotchas, the pitfalls, the lessons learned, if you will?

Scott Blandford [00:43:51]:
Well, when we started our automated assistant project, and maybe I’ll start there, we weren’t sure what topic areas customers would want to use it for. So the way we set ourselves up was let’s start small, let’s launch our assistant, and when we get started, it won’t be that smart. It’ll only be able to answer a very narrow range of questions. The one way to deploy something like this is you try to teach it everything at TIAA, which would take a very long time. They try to teach it everything about us, and then launch. Well, that would have taken years in development.

Scott Blandford [00:44:28]:
We decided to go very small. Let’s launch with a real limited knowledge set, and let’s see what customers ask. If we can’t help them with the automated assistant, we’ll just bounce them over to a real person, but we’ll learn what they prefer to ask through a conversational UI versus a graphical UI. We thought this approach would work, and it did. We’ve got a team of people that basically look at the questions asked and answered every day. As we’ve scaled up where we promote the assistant, we’ve invested in teaching the assistant more of our platform. So I look at it like we have two dials that we’re managing every day. There’s a dial that how smart have we made our assistant? And there’s a dial that how many places are we offering it to customers? And we want to sort of move both dials at the same time and grow our way

Scott Blandford [00:45:33]:
into the platform. But that was kind of a different approach here than we’d used for launching other services. So that was interesting and just bringing everybody along without how we’re going to do this, and we believe it’ll work. The good news is, as it was a new way of launching a platform, but in hindsight, I think we’re very glad we did that way because we learned a lot, kept our investment focus, and we know what people are looking for help with.

Sanjog Aul [00:46:10]:
Again, on behalf of the show and our listeners, thanks so much, Scott, for sharing your insights on the very rationale on why someone should think about AI and ML, and even though we are dealing with the COVID 19, which is a healthcare crisis and is spawning a financial crisis, how do we justify continued investment. Thank you so much.

Scott Blandford [00:46:33]:
No Sanjog, thank you so much for having me. It was a pleasure to be here.

Sanjog Aul [00:46:37]:
Thank you again once. So listeners, please find us on our podcast on Apple, Google, Iheart, Spotify and many more. Connect with us on LinkedIn, Twitter and Facebook and Pinterest. Thank you once again for listening to CTN. You guys should all stay safe. This is Sanjog Aul, signing off till next week. Take care and God bless.