Ep 52: Modernizing a 100-Year-Old Institution: AI at TIAA with SEVP, Chief Operating, Information & Digital Officer Sastry Durvasula

Evan Reiser: Hi there, and welcome to Enterprise AI Innovators, a show where top technology executives share how AI is transforming the enterprise. In each episode, guests uncover the real-world applications of AI, from improving products and optimizing operations to redefining the customer experience.  I'm Evan Reiser, the founder and CEO of Abnormal AI.

Saam Motamedi: And I'm Saam Motamedi, a general partner at Greylock Partners.

Evan: Today on the show, we’re bringing you a conversation with Sastry Durvasula, Chief Operating, Information, and Digital Officer at TIAA.

Serving the investment and retirement needs of over 5 million customers, TIAA is a Fortune 500 financial services company with $40 billion of revenue and over $1 trillion in assets under management.

In this conversation, Sastry shares his perspective on how AI enables large-scale transformation within a 100-year-old institution, why everyone should have a secure AI assistant, and the cultural components that will unlock the next wave of enterprise AI adoption. 

Do you mind just giving our audience a bit of a brief background of you and your role and how you got to where you are today?

Sastry Durvasula: Yeah. I lead the technology global operations, shared services, and digital and client experience at TIAA. So these are the... I call them the four legs of the stool. You know, tech, global operations, our digital and client experience, and our shared services, which constitute almost half, or more than half, of our colleagues across the globe at TIAA. So that's the role that I have. And I have the privilege of leading a very large global team. And that's obviously driving the mission of TIAA, which is to deliver lifetime income for millions of Americans. So that's my current role. 

Prior to coming here, I was the head of tech and digital at McKinsey as a partner. And before that, I was at a global insurance broker, Marshall McLennan, where I was leading up digital analytics and data. And then before that, I had various CIO roles at American Express for 15 years. So that's the longest innings. 

So, I guess the way I characterize my craft is I gravitate to 100 year old companies and lead and reimagine transformation opportunities for them. And most of the companies I've worked at are at least 100 years old, except McKinsey, which is 99. So I think that kind of qualifies, yeah.

Evan: Well, Sastry, obviously, there are millions, at least millions, of people that know what TIAA is. Can you share a little bit more for maybe some of the people in the audience that are less familiar, about what you guys do and how you help your clients and customers?

Sastry: Yeah, no, TIAA, for those of you who don't know, was founded in 1918 by Andrew Carnegie. He founded this company because he realized that teachers and healthcare workers didn't have pension plans. So, it was more to serve those who serve others; that was the mission on which TIAA was built. 

So, fast forward 105 years; we are a pioneer in the retirement space. So, we are like the OG when it comes to the retirement plans for higher ed, for healthcare, and for the nonprofit sector. So, that's our main business. We have a large asset management business. It's called Nuveen, and we manage around $1.4 trillion in AUM assets under management. So, that's our second large business. 

And then we have vault management, which is helping clients achieve their vault objectives. So, it's a B2C and B2B2C. So, the first one is more B2B, and then retirement because we work with a lot of these institutions, but we help their participants. And then asset management is also a pretty large institution, and there is a B2B2C segment as well.

So I think it's a global company. We're headquartered in New York, and most of the C-suite—almost more than half of the C-suite—have joined over the last four years or have assumed different roles since Tashana Brown Duckett joined as CEO. She kind of built the executive team that I'm part of.

Evan: So, you mentioned a minute ago that you have a history of joining hundred-year-old companies to help with transformation, right? And, obviously, sitting here in 2025, it's increasingly clear AI is going to play a huge role in not just digitizing but fully transforming businesses.

Where do you think we're at in the AI hype cycle? Right? I don't think there's many technologies that have ever been as hyped as AI, but I think there's probably a little bit more real stuff there that may be like some past fads.

Yeah, where do think we are in that hype cycle?

Sastry: We're still in an "a lot of sizzle and less stick" place. You know, there's no other way of explaining it, Evan, because we are in the early stages of impact, but we have overachieved on the hype side for sure. 

I think there is obviously enormous potential. There's no question about it. And I think the gravity of the situation is more, what I would consider, the edge cases or sizzle cases than real problems in real businesses that should be solved for that to have scale and, you know, things that can be sustained over time. 

So, I think there is a lot of technological innovation—there's no question about it. And then there's definitely a lot of innovation in the models. There's a lot of innovation in how we can integrate all these good things. But when it comes to real problem-solving for businesses, I think we still are in the early stages as an industry overall.

Evan: One thing I hear from a lot of CIOs I talk to is, they say, well, you know, AI is super interesting, for sure. It's going to be a thing. You know, I like ChatGPT, you know, I see my kids using it for their homework, but I'm not quite sure where that actually applies in actual business, and is it more of a toy, or is it actually going to have a real impact and change things? 

You know, I think, given your background and your current work today, are there areas where you've already seen an applied impact that maybe would give some more hope, or optimism, to some of the skeptics?

Sastry: So I have this 80-20 rule, Evan, that I keep talking about when it comes to AI. And my belief is that 80% of the jobs in the world will have at least 20% impact by AI and that 20% of the jobs will have as much as 80% impact by AI. So, of course, we're still going, and this is not just the technical jobs, or data science, or some of these cool AI problem solvers. This could be anybody in any function. 

So, I guess, with that context in mind, there are really interesting problems that are being solved for. I'll give you a couple of examples within our own environment. We launched our TIAA AI platform, and originally, the OG of that was pre-ChatGPT. So we started with that. When I joined in 2022, on Valentine's Day, it was with lots of love for TIAA. The first partnership we launched was actually with Google in Google AI for call centers. So, we started implementing Google AI in various call center operations. Then Q4 happened, ChatGPT was born, and then the world is different. 

We have since partnered with a lot of different firms and a lot of different solutions. We launched this platform called GATE, TIAA GATE, for generative AI technology. And that's what we've been putting in the hands of our colleagues. 

So, a couple of good use cases. One is research. So, in our asset management space, obviously, we invest in a lot of different asset classes all the way from equities to fixed and real assets to alternatives, et cetera. So there is a lot of research that our analysts and our portfolio teams do. And a lot of that is manual plus semi-digital. But now we have created, with our GATE platform, a research buddy, which is essentially a digital AI-powered assistant that does a lot of this work.

And now it's like scanning through so many documents, but not just scanning through them, but it actually synthesizes them in a generative style, where you could get the snippets that you're looking for, just like how you do it in ChatGPT. So, I think that's a powerful use case because it saves so much time for portfolio management teams that are managing money so that they can actually focus on the money and strategic decisions. They're just doing a lot of chores and getting insight. So, that's a good example of implementation.

Same thing with our retirement business. As I said, we have contracts from a long time ago in the galaxy. So, we have to analyze those contracts and answers.

Evan: You said they're on paper? Is that where you print out a Docusign, or how does that work?

Sastry: Yeah, we have digitized all those paper contracts, but imagine you're calling our customer service, and they have to go through these contracts to answer the questions. It's pretty complicated when you have multiple contracts with us, and you may have changed jobs multiple times from different universities or different healthcare firms where you worked. And now we have to answer the questions because, at the end of the day, Evan is one Evan, but we may have many contracts from many universities or healthcare systems that you were part of.

So, to answer the question, now imagine AI applied to it, especially Generative AI, where all these documents are securely synthesized, and the customer care professional is getting input on their desktop that is so customized and tailored. 

Now you could say, 'Wasn't that possible before?' It was possible to study it. It was definitely not possible to synthesize it in the English-like fashion that generative AI technology offers, right? So, that's actually a great example. We have a couple of those real examples that we've launched in production now. 

We also have some sizzle use cases, as well, that we started with. Actually, it's a good use case, but it's more sizzlish than stakish. We did all the phishing campaigns that we do in the company, like for cybersecurity. We said, you know what, we'll just go away from all these static emails that we used to send, and we would just use Generative AI.

Evan: Yeah, super smart, right? Synthetic is actually a real example now.

Sastry: And these are, I'll tell you, man, these are crazy emails that are so customized, and you would be so inclined to click on the link. Emails that Generative AI craft. Yeah, I mean, it's not like a business impact, but it is actually training our colleagues to a level where this is what the bad actors are using, as well, right? 

So, we are training ourselves up for cybersecurity, and we put this on our TIAA website, as well. So, if you actually go to the website and search for something, there's a lot of financial jargon, like, you say, 'What's RMD, requirement distribution?' and then, 'What's the difference between traditional and raw tire?'

We have traditional search results from a classic search, but then, next to that, we have gate-powered search results. And you can see for yourself the difference as a client, right? As a prospect, I think the opportunities are there. We're still in the very early stages. I am optimistic.

There's also a lot of ethical and governance concerns that we can talk about, but obviously, as a regulator of industry, as a financial services firm, we are very, very, very diligent about it. But I do think that the more we solve these real business problems in these institutions that are iconic and centuries old, I think the more it'll become really mainstream like the other technologies that are mainstream now. This digital collaboration that we are on. You’re here in Las Vegas, and I'm in Charlotte, North Carolina, and we are having a very meaningful conversation. That was not possible like 15 years ago.

Evan: Do you see a world where every worker has a kind of AI buddy or maybe an AI intern that's doing some of the more mundane work that allows people to focus on things that humans are best at?

Sastry: Yeah, in fact, I like the way you said it. Every worker has an AI buddy because that is my ambition for TIAA. We started TIAA GATE, and we said, 'Okay, we're to give you a gate.' And myGATE is something that every colleague in the company will eventually have. In fact, we started with 350 colleagues as beta and they're super loving it. So, now we want to get to around 5,000 by summer, maybe 8,000. My team is going to hate me for saying this on a podcast, but 8,000 by the end of this year. Colleagues having myGATE as their personal assistant is great because I know my workload better than anybody because that is what I do. You know your workload. So, how cool would it be to have your own AI assistant that has all these great large language models propelling it in the backend, but you're removed from all that complexity? You have your own assistant. You can just work with that assistant, train the assistant, be with the assistant. You can even name it whatever you want your myGATE to be. That's going to be very powerful, right? I mean, you could actually offload some of the work that you do to the assistant, and then you could focus more on strategic work or learning opportunities to create a bigger impact, et cetera.

But just applying it now, broadly, I think the bigger impact is what we are not seeing yet. And this could be a bit controversial to say, but I feel that it's leveling the playing field. Imagine the future three, four years from now. You have these startups that are coming up that obviously cannot employ a lot of people because they just don't have the investments, but they can employ a lot of agents that are digital in their workforce because that's easier to do because it's technology. So, now you have companies that have more agents and less humans competing with companies that have lot more humans and some agents. 

Now, how that's going to evolve and what that competition would look like will actually tell how the world is frankly going to evolve. Right? Because we are focused on the 'Will AI take over my job?' problem statement today. I think it's an important question, obviously. And I do think that you could almost apply that to any previous industrial revolution. 

When electricity came, everybody who was working on steam engines probably asked the same question. And then there was a Chief Electricity Officer. That was probably the shortest-lived C-suite job ever. They put electricity, and then they were gone. So, you could apply that technology question to anything. But I do think that for AI, the real big question is when the workforce is a human plus AI workforce, how do you win at that point? How do you compete? Who wins in that race? 

Because humans, at the end of the day, have a body clock. Agents don't. AI agents don't; they're 24/7. So, these are the more tough questions to contemplate, and that's what actually makes this a lot more interesting than prior revolutions of technology.

Evan: Yeah. I mean, you have to imagine the best companies in 10 years. It's not 0% humans, and it's not 0% AI. Trying to figure out 'How do we get the best of our world?' is really interesting.

You know, one thing that I personally struggle with as a leader when it comes to AI transformation is actually not the technology piece, but it's the cultural piece. Right? I've seen we have some jobs where people are a little bit resistant. They like the craft and what they're doing. They don't really want to use AI in some areas where it's obviously better. 

How do you think about transforming the culture of the company? The idea that there's some agent you can ask questions to that has better answers than any human colleague. The idea that an AI might give you some advice. You shared the example of the AI buddy, and I was thinking, 'I'm very privileged that I have an executive coach.' But the executive coach maybe watches a handful of my town hall presentations and gives me some feedback. What if every single person in the company had an AI coach that only cared about helping them improve as an employee? That would be a great thing. 

However, I think, one barrier to that is going to be a big change in how we work is having AI give you critical feedback or help you do your job. That's not what we've been doing for the last 20 years. So, how do you think about evolving the culture, getting people comfortable with that, helping people focus on the opportunity, and maybe removing any unnecessary anxieties. Is anything you've seen effective, or is there any advice you would give to other technology leaders out there about how to help the organization move and get the best of the new parts of this new technology world.

Sastry: Yeah, I think my comment earlier, Evan, on the whole COVID-driven digital transformation, was largely cultural, right? Actually, out of necessity. That we had no option, we had to just overnight digitally enable the whole firm, whichever firm you worked at, and we had to collaborate digitally. So, I think, culturally, we had no option at that point; we had to do it.

Now, here, in this situation, we do have flexibility. Every company can choose to decide how they do it, at what pace, and all of that. But at the end of the day, to go back to your thirds, you would have the utopian side and then the dystopian side and then the middle. 

I do think that every company in its culture has these early adopters, and they're excited about technology adoption. They're like, 'I see my own company.' A lot of people fall in that category, and a third, for sure, in AI adoption. And then there's a third with this very skeptical view of, 'Well, is this like Skynet happening? Is this a minority report? We've seen this movie before. It's not gonna actually scale to the level that these geeks are all hyping it about, right?' And then there's the middle. So, I do think that, over time, it will all even out and become mainstream, like we've seen with other technologies.

Culturally, how much access do you give to colleagues to AI in an unintimidating way so that people get comfortable? If we give them these tools, make them learn, and make them comfortable in leveraging it regardless of their craft, I think that's definitely a big cultural step. And then, honestly, giving them access to collaborate more broadly.

In our firm, we started this thing called Guild Network, where we have craft-based guilds. One of the highest subscription guilds is artificial intelligence. We have around 4,000 members in our Guild Network, and the majority of them are in the AI Guild. It doesn't matter what you do in your day job; you have access to this, you can learn, and you can change your craft if you want to, as well. We give guild gigs. So, you could major in your current job, but you could do it in gigs, and then you do a minor. That way, at least the skills you're learning, you can actually apply, and then you could have mobility. So, I think the more we give that level of access to people, culturally, that would be quite useful.

Evan: From a cultural perspective, I love the guild idea, right? Not just for the actual work and training but even to communicate to employees, 'Hey, we care about investing in you.' Right? We care about you learning about new technologies.

I think that's a great signal that every company aspires to send. But sometimes, they don't mean it. Sometimes, they don't; it's not unclear. So, you guys have done a great job building that culture.

And you have to imagine, for the next generation of leaders at TIAA that come up through these programs, they're going to help propagate that and double down. So, I appreciate you sharing that example.

Sastry: No, absolutely. And I think it's also giving them the tools, too, right? Because I'll be honest, I mean, every CIO struggled with this decision. Like, when ChatGPT came, all the early adopters obviously wanted ChatGPT available inside the company. And it has all kinds of security issues, right? And that's why we built GATE and have given them myGATE. Now they can confidently and securely have their ChatGPT-plus level tool, which actually has all kinds of LLMs behind it, that is secure and that has been built with responsible AI codified into it. They're comfortable now leveraging that. 

So, I think the more we can give them the tools, our associates and colleagues, the more I think we will foster citizen innovation in the company, where you've democratized innovation to a level where everybody has a tool, and they can think about what impact they want to make.

Evan: I'm going to ask you to take out your AI crystal ball, which I think you keep somewhere in your office there. If you're looking five, ten years into the future, what do you see as some of the most exciting ways you think AI is going to transform your industry? Specifically things that maybe some of your peers may underestimate or things that other CIOs may consider science fiction. What do you think people are underestimating, and what are you excited about?

Sastry: Yeah, I think, as a fellow Star Wars enthusiast, I'll probably start with the Yoda code for you. That would be, 'AI or ain't. There is no try.' I think every company will have AI in their core. 

I think the biggest difference that I hope and believe AI will make is in the operational aspects of companies. At least these large, old companies, just like how the internet came and digitized a lot of the physical aspects of the world. We don't stand in line with a token in our hand anymore, right? We go onto the phone and get stuff done. So, I do think that some of those core operational aspects that are very consuming, over time, will get AI-augmented. So, that's one. 

There will definitely be better ways of reaching out to your customers and delighting them with your products because it opens up a range of opportunities to reach out. What we're already talking actively about is search enable, like 'Is SEO still a thing?' for instance. There are new market opportunities that will come out through this from a product penetration point of view.

But then, of course, the other big prize would be AI-powered products that we can't even imagine today because we just don't have that level of knowledge. So, if I were to use my Yoda crystal ball, I hope that there will be new products that we can’t think of today. Not technological products but real business products that would come out. Just like how the sharing economy came and the gig economy came, that was not even an economy before. So, I think I hope there are new opportunities like that.

Evan: Alright, unfortunately, we've only got two or three minutes left, so we're gonna do a quick lightning round. I'm looking for your one-tweet response to questions that could possibly be answered in one tweet. Just forgive me in advance, and hopefully, you'll talk to me again in the future.

So, first question is: how do you think companies should measure the success of a CIO?

Sastry: I think, how is technology powering the business growth? How is technology fueling innovation and client impact? And more importantly, how is it strengthening the core of the business? I think those are the three legs of the stool to measure the success of a CIO.

Evan: What advice do you wish someone had told you when you first got into a technology leadership role and first became a technology leader?

Sastry: Luckily, I actually had the advice. We talked about platforms, processes, and people; it's like a classic three P's. One of my mentors said, 'When you take a leadership role in tech, it's all about people, people, people.' So, to me, that's the advice: at the end of the day, we could talk all about AI, but the human element of leadership: there's nothing more important than that in any career. And that applies to technology leadership careers as well.

Evan: What's a book you've read that's had a big impact on you? Whether it's Star Wars or not. I'm just kidding.

Sastry: Definitely Star Wars. I think there are a few books that I really love. The World is Flat is a book that opened my aperture quite a bit early on. That's one. My Life in Full is another book that I really love. Hit Refresh is another book I like. These are obviously CEOs and technologists who wrote the books, but I'm inspired by when you can tell a bold story of an old company and how you refresh things. I think that's my gravity point, anyway, so you can see why I like these books because they have a common thread amongst them.

Evan: What's an upcoming technology that you're personally most excited about?

Sastry: Quantum. I do think that we've talked about it quite a bit, just like the world talked about AI for a long time.

Evan: I mean, I've got some books written here in the fifties that were talking about AI. It's crazy.

Sastry: Yeah, all the way going back to the Turing test, right? So, I think quantum, same thing. Quantum mechanics and quantum physics, we've talked about it for a long time. I do think that the impact is still awaited. 

And, of course, there's progress to be made. I was just reading that there's a 75-qubit quantum computer that's being used to train models in China. I mean, imagine putting the power of quantum and AI together. And you're training these models and unleashing them. Or Majorana, which Microsoft obviously released, is more of a topical example. So, I think quantum is what I'm more excited about. I'm actually, personally, in my own learning journey, spending a lot more time just studying the possibilities of quantum and when the quantum rises.

Evan: Okay, last question. We'll stick to the science fiction thread here: what do you think is going to be true about AI's future impact on the world, beyond just business, that most people would consider science fiction today?

Sastry: I hope we won't have flying cars anytime soon, like in the science fiction movies. This is where I really hope that AI actually helps our day-to-day chores. How cool would it be to really have robotic helpers beyond digital assistants?

My wife, Debbie, always says, 'Dude, all the work you do in AI. Can you actually do this? So that we can get more time for ourselves?' I hope that it actually helps the good of society. Hopefully, it releases more time from people so that they can actually build a better human connection. That's my hope, but then, of course, it could also become like the cell phone, where family dinners are happening with kids on the cell phone, not even making eye contact. So, I think it's more to be seen on the silver screen.

Evan: Well, Sastry, it's so great to see you. Thank you for making time. Hopefully, we can chat again soon.

Sastry: Bless you. Thanks for having me, Evan.

Evan: That was Sastry Durvasula, Chief Operating, Information, and Digital Officer at TIAA.

Saam: Thanks for listening to Enterprise AI Innovators. I’m Saam Motamedi, a general partner at Greylock Partners.

Evan: And I’m Evan Reiser, the CEO and founder of Abnormal AI. Please be sure to subscribe so you never miss an episode. You can find more great insights on enterprise AI transformation at enterprisesoftware.blog.

Saam: This show is produced by Josh Meer. See you next time!