Audio:

From the New York Stock Exchange at the corner of Wall and Broad Streets in New York City, welcome Inside the ICE House. Our podcast from Intercontinental Exchange is your go-to for the latest on markets, leadership, vision, and business. For over 230 years, the NYSE has been the beating heart of global growth. Each week we bring you inspiring stories of innovators, job creators, and the movers and shakers of capitalism here at the NYSE and ICE Exchanges around the world. Now, let's go Inside the ICE House. Here's your host, Lance Glinn.

Lance Glinn:

Artificial intelligence is fundamentally reshaping the way we work, transforming nearly every industry and profession. From streamlining operations to powering smarter decision-making, AI is now an essential part of the modern workplace. It's not only boosting productivity, it's redefining job roles, skill sets, and the pace of innovation itself.

At the forefront of this shift is Writer, a company empowering enterprises to build and manage custom AI agents that automate complex, time-consuming tasks across departments. Kevin Chung, the company's chief strategy officer and our final guest in this special three-part AI series, is helping Writer unify teams and drive enterprise-wide transformation. Kev, thanks so much for joining us Inside the ICE House.

Kevin Chung:

Yeah, thank you, Lance. I'm excited to be here and chat with you.

Lance Glinn:

So Kev, before Writer became the leader in enterprise generative AI that it is today, May Habib and Waseem Alshikh saw an opportunity to redefine how people work. It was in 2020 that Writer was launched. You joined the company in 2023 as COO before transitioning to your current role in 2024. Just what about the leadership of May and Waseem and the platform as a whole convinced you that this opportunity was the right one for you?

Kevin Chung:

Yeah, I mean, I've been a part of a number of pretty large digital transformations. I met May in early 2023. And this was at the time where I think we had just coming off the pandemic, we'd just come off everything that was pretty hyped around, I would say, metaverse and there was a lot of things going on. And so people were getting really interested in AI. And so I was looking at it and saying, "Hey, are we hitting another hype cycle? Are we going to really see something really interesting happen here or not with AI?"

And when I met with May and Waseem, immediately what I had realized about them was their determination and drive of just owning and winning a technology space in the enterprise was very, very key. So I've worked in a number of companies where there's been a really, really strong focus on enterprise and there's just so much dedication has to go into it. And I remember my first meeting with May, she was just showing me all these accounts.

She was like, "These are all the CIOs and CEOs and all these executives we're talking to." And I was just immediately so impressed by how deep of knowledge that she and Waseem had and how much dedication they had there. And so a big part of it was really, one, you obviously have to have a great product with great product market fit, but you have to have really great leaders and visionaries who are like, hey, we are in this business.

We understand them. But not only that, we understand what our customers really need and want. And so that really sold me is just their deep, intimate knowledge of understanding the customer, but not only the customer, the executives behind those customers and what they're doing. So immediately I think within five minutes of meeting May, I was like, wow, this is a CEO who's really going to take this company into some interesting places. I was really just drawn by their attentiveness to the whole space and just how they were dealing with it.

Lance Glinn:

And you mentioned some of those customers, NYSE-listed enterprises including Salesforce, Uber, Kenvue, Lennar, among many others that are globally recognized companies that trust Writer to maximize productivity and creativity. Just what about the platform and what the team provides to clients allows Writer to really separate itself in such a competitive space?

Kevin Chung:

Well, it's pretty crazy is in a short amount of time the amount of education that we have done to bring the market up to speed on this. I mean, it really started with building a platform that was, number one, enterprise-grade. That was super important. Number two was one that you can inspect, observe, and see everything top to bottom. We're not only talking just like LLMs, but the ability to build apps, the ability to use your internal data, the ability to manage your brand, tone, and voice, and do all of those things.

I think having the simplicity of having all of that built on one solution without you having to pick and choose and figure out, I'm going to use this and does it work together, and is it compatible. That was one really easy thing for enterprises to say, "Look, I have an architecture that has worked for hundreds of enterprise customers and thousands of workflows already today," was a pretty key thing.

But I think the big part about it was actually what we call the last mile was to get all of it to work and then to be able to scale it. Most customers aren't talking at this point about one or even dozens of use cases. They're talking about hundreds of thousands of use cases, workflows and agents that they're trying to completely transform their business. And you can't build that scale if you're looking at it at one singular use case or one platform that doesn't allow you to go across departments, across different use cases.

So a big part of it was, one, the architecture really just has to be compatible and work within enterprise. But then two is like how do you actually scale this to your marketing department, your product teams, your people team, your R&D, all those types of things? How can you have a platform that allows you to do that? Because you don't want to buy point solutions for every single one and then have to train it.

That's just so much time and effort. So I think the ability, one, of just having the architecture work within your environment, but two, to be able to scale it to at this point thousands of tens of thousands of different workflows was really the key differentiation for us to be able to scale in the enterprise and take that leadership position.

Lance Glinn:

So I want to pivot now to some recent developments for Writer beginning with AI HQ, a product was launched in April and designed for enterprises to orchestrate agent powered work. Now, you just spoke to obviously creating a platform or having a platform that fits many, not just one specific unit. So how does a platform like this, like AI HQ, bridge the gap between IT and business teams and what challenges is it helping enterprises overcome in their adoption of GenAI?

Kevin Chung:

Yeah, absolutely. I kind of view AI HQ as an evolution of what we've been doing for the last few years, and we have been actually uniting business and IT for quite some time. A big part of it was the use cases people were solving I would say in 2020 really just like, here's a part of a workflow or things we want to augment and make it just AI ready. So we were doing a lot of things that were part of a workflow.

And what AI HQ is now allowing us to do is orchestrate almost fully autonomous end-to-end workflows. So the uniting of business and it has actually always been a staple of what we've been doing and helping us do that and pieces of workflows within each of those groups. But now with AI HQ, it's actually allowing you to orchestrate the entire three autonomously across end-to-end.

A really great example might be imagine recruiting. We're trying to hire a bunch of new people to Writer. We're trying to hire hundreds of people before the end of the year. You can imagine the number of people in an organization involved in hiring. You've got to hire them. You've got to summarize all the feedback from people. You've got to actually score them. You have to be able to issue offer letters.

Legal has to generate contracts. You have to be able to review those with the compensation team. So all those are individual steps within an entire workflow just to hire one person. So now with AI HQ, an example of us building multiple different agents to manage the recruiting process, you can have an agent that summarizes a candidate's entire feedback, one that will analyze it against a set of requirements of how you want to score them.

Then you can also have an agent that's going to actually issue you an offer letter with the exact terms and salary and benefits that you want to give them. And then you can also have another agent that can generate the welcome letter or the onboarding documents, all those different things. And originally even before AI was like a human that had to be in an individual application and creating that content, now you can have agents actually initiate and orchestrate that whole end-to-end workflow.

So now you've got the people team, the legal team, the hiring teams, the account teams that you might be hiring in a year, are all part of this process and all able to do that at one time. And that's just one agent of thousands of agents that we've already built already that you can actually start to build all these things together. So really the evolution is really allowing us to take more and more and allowing AI essentially and AI HQ to manage that whole process. So it's really the build, supervise, and orchestrate the whole thing altogether.

Lance Glinn:

So later in April, Writer released Palmyra X5, its new large language model. Palmyra boasts a groundbreaking 1 million token context window and is setting a new baseline for enterprise AI. Can you break down just why that matters and what it enables for real world use cases?

Kevin Chung:

Yeah, absolutely. So a big part of it is you want to use your own internal data, and your own internal data is the reason you're able to get really reliable and accurate responses to anything you're doing, whether it's a Q&A that you're trying to create or a chatbot you're trying to create around Q&A, or you're doing deep analysis on a market sentiment as a financial institution.

So a 1 million context window, so just to give you what that means in terms of size, that's six large books or 1,500 pages. You can put all of that context into something that you're trying to analyze and be able to get a response back to it very quickly. So imagine being able to actually condense all that information, that volume of information, and basically say, "Hey, I need to either analyze this, summarize this, research this, use all this context to create something with AI."

And not only that, you can do that in a matter of seconds. It's just the speed of which you can do it is tremendous. It's like we're processing a million tokens in 22 seconds. That's just an incredible amount of processing that you can do very quickly. And so what's really phenomenal about that is it's just we've advanced so quickly in how much that we can do and use AI for.

Imagine if you're doing research as a pharmaceutical company and you're literally putting decades of your research work in there and trying to come up with, hey, help me put up together a thought leadership piece or help me actually summarize this decade of work that I've been putting together in my company's own tone of voice so that it's understandable to a marketing audience or a audience of researchers.

Being able to do that in a matter of seconds is really, really powerful, where it used to maybe take somebody a month or a week or however long to actually do that manually. And so that's the importance of how much you can use these large context windows, but also do it in a very cost-effective way. We've heard a lot of these stories where you're running these other use cases or models where it's super, super expensive.

You need to also be able to do it efficiently. And so I think that's the balance of what we see with Palmyra is you're able to do things highly efficiently, but also get the results that you want in the leader in everything that's going on in the market right now.

Lance Glinn:

So AI HQ, Palmyra X5, these innovations I think speak to the ever evolving nature of AI as a whole. It's obviously a technology that's changing all the time, and the role that companies must play in ensuring that that technology acts responsibly and without bias. So how is Writer approaching that challenge, ensuring that clients have trust in what they're using?

Kevin Chung:

Yeah, absolutely. I mean, a big part of earning that trust is working really closely with the enterprise and understanding what their requirements are in terms of what is specifically they're trying to solve for, the workflows and the output. So one part of it is actually really close partnership with enterprises and really deeply understanding their workflows, their business, and just doing that as a partnership.

So that's number one. But number two is actually the technology too. And even when I started in 2023 and as I talked with May and we've seen the things that they were hearing in 2020, a lot of enterprises were concerned about hallucination. What will the AI create and is it something that is going to be accurate or reliable? And the one thing that's really uniquely different in AI and enterprise versus consumer is you have to meet a really, really high bar.

But the other piece of it is when you're building an architecture that is actually relying primarily on customer's internal data, you're able to benchmark that against what is the response and accuracy that you need. We're not pulling things from the internet or pulling things from things that are just not part of the day-to-day workflows. And so that is a big part of it. So I guess what I would really say is when you think about working hand in hand with enterprises is really owning their trust.

That partnership is super key. But then on the technology side is really like how do you actually rely and use their internal data to actually get the accurate and reliable responses you want? And that's the reason why when we were talking about X5 earlier, is being able to have a technology that can really analyze, compute, and do things at scale with a large amount of documents in an incredible speed and amount of time is our ability to get there.

And so nowadays, there's actually no conversation we have about concern about reliability or hallucination or anything. It's more about how do you actually allow our business to move faster, and that's a great place for us to be in.

Lance Glinn:

So I mentioned earlier in our conversation, started with Writer in 2023 as COO, transitioned to chief strategy officer in 2024. As chief strategy officer working with May and Waseem and the rest of the leadership team, how do you mesh near-term business growth opportunities while at the same time, especially with the technology like AI, looking long-term and looking at innovation in such a rapidly evolving landscape?

Kevin Chung:

What I would tell you right now is what we're seeing at least, there's obviously the short-term of we want to enable and help as many enterprises as possible on their immediate term goals, but what we also view is this technology changes so fast. Almost every couple of weeks we learn about another use case or there's different things that enterprises are trying to do.

And what's really exciting about where we are right now is I actually think that the technology that we're building is outpacing right now what enterprises want to do, which is great. And here's the reason why, because we're able to together educate the market and with our customers on this is the way that you used to do it. Imagine a world where you can do it a thousand times faster, a thousand times more accurate.

What's really exciting about that world is it's going to completely transform the way we do business. It's like one of those really phenomenal step changes that we're going to see in terms of being able to reinvent the way we work in healthcare, financial services like retail, even in big tech. So that's the part that's super exciting.

A big part of it is, yeah, we're inherently in the moment right now with our customers being like, hey, we're going to solve this problem for you, but we're also as the major enterprise research lab on AI thinking, well, we need to think five, 10 years ahead to be like, what problems can we transform? What evolution of the product can we build that's going to help us get to that next level?

It's already really exciting what we can do today, but even in six months from now, there's going to be some tremendous things where we're completely changing the way a business rethink responding to RFPs or giving your clients updates on their portfolios or even how we think about R&D research for pharmaceutical drugs. It's incredible to think that even six months from now, we're going to completely revolutionize the way that's being done in-house for every company.

Lance Glinn:

So we're looking, you say, six months from now. When you think though about the next five to 10 years, even looking further out than say 2025, '26, 2030, 2035, et cetera, what does the future of AI look to you, especially as it obviously moves deeper into how organizations think, create, and operate?

Kevin Chung:

Yeah, so we've always laughed about it. And I think you've seen this in media where it's like there's a future where we're working side by side with robots. And it's like the AI evolution is really not that different where we'll obviously have human coworkers, we'll have AI agents who are actually working for us.

This isn't a five-year thing, but what I would say is the steps to get there and what we might see a decade from now is there's a lot of things in our day-to-day work that are highly manual, that requires a lot of compute and may not require now a lot of human compute to be able to do that. It's like we'll have AI be able to handle that for us.

And so I see the world even just like a year, but if you say even five or 10 years from now, where we have really, really sophisticated and knowledgeable humans who are able to orchestrate working with agents and leveraging those agents to do so much more. So example might be, look, my day-to-day might be I come in and I open my emails. I'm looking at my Teams or Slack. I have a bunch of things I have to respond to, and then I have to create a bunch of content.

Well, now if future five years from now, I'm able to do all of those things in the first one minute of my day, where it's like it summarized the most important things I need to do. It's actually started to create this content, but I'm actually the conductor behind the scenes being able to orchestrate where I spend my time, does that output actually meet the standards that I want and allow me to actually think more strategically on the bigger things I want to do.

So instead of being able to accomplish what I need to do in one week's time, imagine I'm done with all of that before breakfast. And now I'm moving on to the next thing and trying to figure out what that looks like. So that's probably a thing we're probably a year away from. Five years from now, a big part of what's really exciting about this is we learn with our customers around where people want to take the technology.

We're just building the I always call it the rails for everybody to get there. So five years from now, I would just say it's going to be a completely different landscape in business and just people how people work.

Lance Glinn:

Absolutely. Now, Kevin, as we wrap up our conversation, so we obviously spoke to major advancements, AI HQ, Palmyra X5, earlier in our conversation. But just under the direction of May and Waseem, how is the leadership team of Writer just plotting the company's future to build on the already great success it's achieved and carry it forward to the years to come?

Kevin Chung:

I mean, we have really, really big aspirations. I mean, when I was mentioning if you think about a mission, what is the mission of the company, we really want to be able to transform every enterprise and help them bring them into the AI evolution. And that's a big part of the overall mission. So we're in hundreds of enterprises now covering Fortune 10 to the Global 2000. We want to run the Global 2000. We want to be the enterprise platform that is running the enterprise.

And so that's the big strategic thing that May, Waseem and I and the entire leadership team are looking at this and saying, "We have the technology, we have the enterprise leadership position, and we have all these great partners that are helping us get there. We have tens of thousands of use cases that we're solving right now. The big thing is how do we scale to help an enterprise wall to wall? How do we go and help every single Global 2000 enterprise help them do what they need to do?"

Everything from RFP responses, research and development, financial reporting, legal documents, insurance claim analysis, medical record analysis, all those types of things. Our big thing is how do we ultimately help scale to as many enterprises as possible across as many different use cases? Because as I was mentioning to you earlier, being able to actually solve all these things for every single one of your employees to help you really get that scale is really the phenomenal thing we're going to start to see.

And just like when we have this crystal ball and go back and look back at it just today and say, "I'm surprised that we were doing these things today." 10 years from now, it's going to be like, "Hey, AI was this really, really powerful tool we wish we had uncovered sooner." And that's the world we're living in, and that's the vision that we see ourselves is helping enterprises and everybody get to that state as soon as possible.

Lance Glinn:

Well, Kev, I appreciate and find it so insightful what May was seeing, yourself and the whole Writer team are doing, and I thank you so much for joining me Inside the ICE House.

Kevin Chung:

Yeah, I appreciate you, Lance. Thanks so much for the conversation.

Audio:

That's our conversation for this week. Remember to rate, review, and subscribe wherever you listen, and follow us on X @icehousepodcast. From the New York Stock Exchange, we'll talk to you again next week Inside the ICE House.

Information contained in this podcast was obtained in part from publicly available sources and not independently verified. Neither ICE nor its affiliates make any representations or warranties, expressed or implied, as to the accuracy or completeness of the information, and do not sponsor, approve, or endorse any of the content herein, all of which is presented solely for informational and educational purposes.

Nothing herein constitutes an offer to sell, a solicitation of an offer to buy any security, or a recommendation of any security or trading practice. Some portions of the preceding conversation may have been edited for the purpose of length or clarity.