From Data Silos to Intelligent Platform: Nasdaq's Enterprise AI Strategy

Summary

Nasdaq unified data siloed across multiple business units by adopting a centralized medallion architecture with Unity Catalog governance on Databricks, creating a single source of truth across 130 global markets and eliminating duplicate ingestion pipelines.
The transformation delivered five times faster index operations and positioned Nasdaq's data as AI-ready, enabling new AI-powered products including Verafin for fraud detection and eInvestment for AI-assisted investment decision support.
Nasdaq's experience demonstrates that enterprise AI strategy must be grounded in a strong data foundation, with centralized governance and reproducibility for audit compliance enabling both internal efficiency and innovative customer products.

From Data Silos to Intelligent Platform: Nasdaq's Enterprise AI Strategy

Financial institutions face unprecedented challenges unifying disparate data sources while maintaining regulatory compliance and governance. Nasdaq, as a trusted financial infrastructure provider, processes exchange data, index information, and market intelligence across 130 markets globally. Data siloed across multiple business units limited visibility and created inefficiencies, preventing the company from scaling innovation. The challenge required a unified data platform to consolidate information, ensure reproducibility for audit compliance, and enable faster product innovation.
Discover how Nasdaq transformed its data landscape using Databricks. By adopting a centralized medallion architecture with Unity Catalog governance, the company unified siloed data across multiple business units and eliminated duplicate ingestion pipelines. The corporate data lakehouse created a single source of truth for product data, sales information, and market intelligence. The result: five times faster index operations, AI-ready data enabling new investment products, and AI-powered applications like Verafin fraud detection and eInvestment decision support, all powered by Databricks Genie and AI functions.
🤝

Chapters

FAQs

How did Nasdaq use Databricks to unify data across business units?

Nasdaq adopted a centralized medallion architecture on Databricks with Unity Catalog governance to eliminate siloed data and duplicate ingestion pipelines across multiple business units. The resulting corporate data lakehouse serves as a single source of truth for product data, sales information, and market intelligence across 130 global markets.

What is Verafin and how does Databricks power it?

Verafin is Nasdaq's fraud detection application built on AI-ready data from the Databricks Data and AI platform. The unified data foundation and AI capabilities available through Databricks enable Verafin to analyze patterns across financial transactions to identify potential fraud for market participants.

How does Nasdaq use the eInvestment platform to support investment decisions?

Nasdaq's eInvestment platform uses AI-ready data from the Databricks lakehouse alongside Genie and AI functions to provide investment decision support for clients. The platform enables users to query and analyze market intelligence conversationally, helping them make better-informed investment decisions based on Nasdaq's data assets.

What performance improvements did Nasdaq achieve after migrating to Databricks?

After adopting Databricks, Nasdaq achieved five times faster index operations, enabling more responsive data products for global markets. The team also eliminated duplicate ingestion pipelines, reduced manual data work, and built a reproducible and audit-ready data foundation required for their operations as a regulated financial institution.

Full transcript

[00:08] Good afternoon. Hello, and thank you for joining us today. Today, we'd like to take you through Nasdaq's data journey, and how our partnership with Databricks has enabled us to make better decisions, drive better insights into our business, and fuel growth for our products and our customers.
[00:24] In this session, over the course of the next half hour or so, we hope to describe how it enterprise data and AI come together in the real world. Using examples and case studies drawn from our experience as a regulated financial institution.
[00:39] We'll leverage and talk about how we think about data as a key business enabler, and how Databricks technology has helped us to manage our entire data estate. But, it's more than about tech. We'll talk about how data, people, and platforms come together to drive the
[00:57] kind of impact that's necessary for true enterprise scale. And because no talk in 2026 would be complete without a discussion of AI, we'll also discuss how we see the data strategy as being integral to AI strategy, and how we're building off of
[01:12] our strong foundations in data and Databricks technology to build innovative solutions for generative AI, both to power our own business, and also to deliver on new products and new services for our customers. My name is Edwin Aoki, as you heard, and
[01:28] I'm a senior vice president in global technology at Nasdaq, responsible for developer experiences and data strategy and governance. And I'm joined here by my friend and colleague Angie Ron. And I've been introduced, but I wanted to double-click on the division uh that
[01:45] I'm in, Capital Access Platform. So, we are responsible for listings, Nasdaq index, Nasdaq data, and workflow inside of many product. Our vision is to bring the capital to opportunities. Well, before we getting started, we
[02:01] actually have some fun activities. And just pay attention because I have something later on for this. So, there going to be trivia question. Of course, I don't expect you to know Aden or Angie.
[02:17] But I expect you to know Ali Ghazi. CEO of Databricks. Oh, there's actually one thing common for all of us. But without telling you exactly what it is, it's something related to the theme of the conference. But at the end of the
[02:34] session, we're going to pick three people. We have some gift Nasdaq gift to share with you. So, now pay attention to the rest of the conversation. Now you're going to have everybody thinking about this during our entire talk, Angie. You can catch the audience, I'm sure.
[02:49] good. Well, listen, hopefully some of you have heard about Nasdaq. Yes, maybe? Um from its start over 50 years ago as the first fully automated trading system, you probably know about Nasdaq as the home of where uh the world's leading startups,
[03:05] innovators, and technologies come to access capital markets and access liquidity. In fact, we're very proud that over 5,600 companies in the US, the Nordics, and Europe have chosen us as the place where they want to list their stocks,
[03:21] including nine of the 10 largest public companies. But there's a little company last week that actually listed on Nasdaq. What was that? I think a rocket company, SpaceX. That's right. That's not part of the quiz. So, no swag for that. Thank you. almost mispronounced big and small.
[03:36] But Nasdaq is a lot more than just listings, right? Right. So, when I first joined Nasdaq many years ago, I thought Nasdaq was like just a stock exchange. Not only we operate 19 stock exchange, US, Nordic, and Baltic, but it's
[03:52] actually a tech company. Our platform powers more than 130 markets and and and regulators in the world. Think about the market. Think about a country, a market. Uh Mex- Mexico. Okay, yes. We power that country, uh the market. What's another one? Uh Japan.
[04:08] Okay, yes. Well, just think about 130 markets around the world. They're powered by Nasdaq technologies, not just the trading, risk management, anti-financial crime, etc., etc. Not only that, you have about listings, Nasdaq index, Nasdaq data, and many
[04:23] things like that. That's right. So, our mission is to be the trusted fabric of the financial system in the world. There's one secret. The one of the biggest strengths of Nasdaq is the core of our
[04:39] gold standard data. Gold standard data. That's right. Our clients who are banks and brokerages, uh regulators, the markets that you mentioned, and companies of all shapes and sizes trust Nasdaq to build performant, well-engineered solutions for their hardest business problems. But
[04:56] none of that would be possible without access to timely, accurate, and actionable data. And that's really where Databricks comes into the equation. Our relationship with Databricks goes back over a decade. We started by using their managed Spark environments to improve deployment consistency and
[05:12] enhance our operational efficiency. We evolved with uh Databricks, or they evolved with us, I should say, to enable the consistent deployment and workflow and production-grade pipelines. And we collaborated with their engineers on new architectural patterns, uh Terraform-based deployments, and
[05:29] innovations like Delta Lake and MLflow to support the needs of larger regulatory organizations like us. And Angie, some of your teams were at the forefront of that. Yeah, absolutely. So, the division of the capital access platform, we started real early. But with the one purpose.
[05:46] We just saw a business problem, a little problem, maybe compliance issues, maybe data quality, maybe efficiencies. But we turn that into solving entire business problem, then the division on the data, then actually bring the value on AI. As a matter of fact, we released
[06:04] the AI ready data and for investment on the Databricks and the data business MCP in the private preview. Very exciting. It is very exciting. And so today, you can look at all the thing we're doing. We actually leverage so many
[06:19] things in Databricks. We only give you a slice of the stories, a slice of story to share with you about some common practices, well, the journey we have gone through and some exciting things. Hopefully, you can take something exciting away and to share
[06:34] with yourself and with your colleague. That's right. So, we're going to take a step back in time a little bit um and start at the beginning of our journey. Like financial markets, many of you are probably in this space, live and breathe on data. And you can actually think of Nasdaq as a data business from
[06:50] the very very beginning. Not only do we rely on that to power our own activities, to be able to fuel the markets, but we also have a robust data business uh ourselves, where we commercialize and monetize that data, alternative data that we use to help the financial services industry get deeper
[07:07] insights and deeper information about the markets. And one of the ways that we've done that and that you may be familiar with is our index business, which you've mentioned earlier, Angie. But people might not know exactly the intricacies of what that is. So, maybe we can help folks understand that a little bit.
[07:23] Yeah, let's go there. Well, just take a look at this graph. And a pop quiz, well, who have invested Nasdaq 100? Raise your hand or in your portfolio. Okay, 10%. Who have heard about Nasdaq
[07:40] 100? Like probably 90%. It is iconic. But it's not just iconic, it's personal. What I meant is that it is biggest conflict between me and my husband. Ooh, what I meant is that 10 years ago,
[07:57] what actually he used our money to invest in S&P 1 500. And if he's investing in actually at Nasdaq 100, you get 600 plus percent and 22% annual returns versus S&P 500, which is
[08:14] over 300 in 10 years and 6 15.6%. I personally prefer 22%. So, whenever I feel I have a fight with my husband, I was like think about that. And this is his biggest regret. So, anyway, I know this is not about my
[08:31] personal story. Let me going back to the real topic. investment advice, by the way. This is also not relationship advice. Thank you. Uh well, let's let's going back to Nasdaq story. So, you started with Nasdaq story, which is
[08:47] 1971. Oh my god, 55 years ago. Nasdaq was born in electronic trading. What that means that Nasdaq never had a trading floor. Just pause a little bit. You see that like like ring the bell, whatever, it's like digital. Like behind the scenes, all
[09:04] digital. But at the same time Nasdaq composite index was born in 1971. Not until in 1985. Think about how many years ago if you invest that Nasdaq 100 was created.
[09:20] It was created out of the most valuable the 100 Nasdaq non-financial company listed in Nasdaq. Okay. The however part Nasdaq index business,
[09:38] not the Nasdaq index 100, actually we use more than just Nasdaq equities and securities or listed on Nasdaq to make indexes. As a matter of fact,
[09:55] Nasdaq index business have more than 10,000 indexes. I'm sure you're going to go back and whatever chat GPT the whatever understand what all the high performance index we have. I'm telling you is imaginary, incredibly exciting. So, out of those indexes,
[10:12] we're not using our own market data. We're using more than 60 markets around the world. And all different variations in terms of the options, in terms of, you know, equities, all the other forms. But, the
[10:30] important thing is that well, we use the market data and the data across other exchanges to do defining the index, calculate a rule-based and transparent leading
[10:45] indexes like Nasdaq 100. So, just the key thing is it's a rule-based, transparent indexes, which is what Nasdaq 100 is based on. And among many thing is that it's really foundation for more than hundreds of ETF
[11:01] that powers more than 1 trillion asset management. Yeah. So, what we do is a lot of technology behind this, things of defining, calculating, a distribution, etc., etc. That's what
[11:16] we do. a lot of data, Angie, and you've applied technology to help address all of that, right? So, maybe we give folks a little bit of an insight into that? Right. So, every single index actually behind the scene, and I know there are actually few people coming index team, so thank you for having me to represent you all. Well, it started with the first
[11:34] thing. Guess it's all around the data. It's incredibly complex system engine behind the scene. The number one is not only we actually ingest, you know, getting our own the data come out our market, we ingest more than 50 vendors, the data come 50 vendors from corporate
[11:51] actions, you know, of course the pricing, as well as foreign transactions, you know, exchange rate, as well as uh fundamental such as companies performance, all of the stuff. And second,
[12:06] as I ASCO, regulatory index administrator, ASCO is really hard to pronounce, it's International Organization of Security Commission Organization. It's must force, like ISO I ASCO. Well, it we demand absolutely accuracy
[12:25] or reproduceable of every single code base from dev to production, every single query test. No No errors is tolerant. What really this mean is that high quality of the data, and we don't want to copy the data multiple
[12:40] times, so shadow copy here and there whenever we access them. And a lot of time, whether it's rebalancing, distribution, dissemination, could happen in the real time. Did you feel like there's some technology can help us to solve this
[12:55] technical problems? I don't know, Angie. We're at a Databricks conference. I smell I taste a Databricks here. So, Databricks really help us in doing many, many things. That's right. In fact, Databricks today is embedded in almost every layer of that index life cycle that Angie was
[13:12] describing. Delta Lake gives us ACID-compliant bi-temporal pipelines, which are really critical for that regulatory audit trail and reproducibility. We use Lake Flow and DLT to power our streaming and batch ETL processes across all of those data sources that Angie was uh mentioning
[13:28] using a Medallion Lakehouse architecture. Our research teams use Databricks SQL to develop new investment theses and manage rich index composition discovery without having to go to a completely different tool chain to do that. We use Unity Catalog to enforce governance, traceability, and visibility
[13:44] across all of that support our IOSCO regulatory compliance needs. It is hard to say. And of course, we have Mosaic AI and MLflow that drive anomaly detection and allow schema detection so that we can catch distribution shifts before they actually propagate downstream into the
[14:01] rest of the pipeline. And the result? Well, as you can see, like incredible business we're running up for index. And one of the things last year uh 76 uh new index was actually created through this new platform.
[14:17] We feel really at the beginning of this, even though it has been many years. So, it's it's incredible to see how technology really enable the index business to transform. Yeah, it sure is. But, Angie, you've been saying that Cap Capital Access Platforms is more than just listings, and it's more than just index. So, we
[14:34] have another you read my script. Yes, yes. your script. Yes, that's Yeah, I do have other other platforms, right? So, maybe we'll go into that. Right. So, as I mentioned, the Capital Access Platform have the listings, index, and and and data business, investment, all etc., etc. Guess what? What you heard
[14:49] about index transformation, it's happening all at the same time. As a matter of fact, data business probably very, very early as well. We all doing this effort. Let's transform our silos here, called silos, in a good way, cuz business ain't go
[15:04] fast, go fast. We're all doing that. Well, the consequence of doing this is that wait a second, are we duplicating the same work? We We actually trying to onboard exactly same data, multiple time, multiple copies.
[15:20] This is expensive. As a technology one thing we like literally a lot of people saying, "Yeah, take a take to the platform. Let's do the sharing." By the way, I don't know if you work in enterprise. When you have the different business unit, when you do the data sharing, it is a really really really
[15:36] easy. Actually, that's not. It's really really hard. It's really engineering project. So, what we have to do throughout like partnership with our product and business is we really have to do some fundamental changes, which is organization.
[15:51] So, the first thing we did, like we actually created a data platform organization. What their job is providing common capabilities for data capabilities. And they provide doing the pipeline instead of you do one pipeline, I do another one for the same data. Wonderful, right? sense. But that's not enough.
[16:08] It's not? Well, why it's not enough? Because we have all different Databricks account. We can't share. So, the actual infrastructure is fragmented. Mhm. We really not be able to share. Like you have to say, "Admin, are you sure you're able to you know, you can't do that, right?" So, we really really partner
[16:24] with Databricks. And you have heard about like catalog catalog. We actually have now migrating to one Databricks account. One catalog. One control. One lineage. The magic is that in the future of
[16:41] sharing is really a governance exercise. It's a click and not a project. So, really proud to see that. So, as you can see the platform transformation is not just technology, by the way. It's It's all the other arts around. The
[16:56] organization and support, the alignment, etc. etc. You know, we call it aligned autonomy. You sort of have the principle, think about the platform, but how do you enable common things where you actually enable all different business unit to go fast, where you do the common things all together?
[17:12] Yeah. That sounds exciting, Angie. How does that How's that worked in the real world? We promised these folks there's some real value. Right. So, at the end I think I really like what Austin was talking about in one of the session is that this technology is all about business transformation. The result
[17:27] number one is of course you can see fast fast fast, more efficient. So, we got a five times faster in data ingestion and index building faster and guiding the private market data, you know, a lot faster. But, what I'm really proud is actually out of the data
[17:43] sharing our index business the research team they actually be able to access securely with the right access right to the private market data. First time ever in an asset we created
[18:00] first private market index. And we launched this in February because of this platform. So, new products, new features faster and entirely new streams of data. And what we see is that making data available is really only the table
[18:15] stakes because really what we're trying to do is make that data more valuable. More valuable for our own businesses, more valuable to create these kinds of new revenue opportunities. But, the really exciting thing that we see is when you are able to take that intelligence embed that into every product, into every new data set that
[18:33] means that you get that multiplicative effect. Every time that you bring data into the platform, it's actually far more than just the sum of its parts. And the speed and in fact the cost of iteration goes down significantly and allows us to move faster and bring new products to market. So, our data journey
[18:49] and Databricks has enabled us to do more with the data. And that's why we say that, you know, a data platform is what makes data available, but it's really the intelligence platform that makes it valuable. And we do that every single day safely and at scale. You, by the way, I we actually had this
[19:06] all this back before even Ali talked about it. He talked about the whole context. You're really ahead of Ali, right? In terms of all you're talking about, you know. We'll see about that. Just trying to connect with him. We'll see about that. But, we are taking those lessons and we're applying that
[19:21] across our enterprise as well. So, you heard about what Angie and the team have been doing across our products. But, really it was the value that we saw in this platform approach that's led us to standardize upon Unity Catalog and Databricks for metadata management across Nasdaq's entire corporate real estate uh data estate as well.
[19:38] We recently completed a 2-year initiative to bring everything together into a corporate data lakehouse so that we can have everything from product data, sales information, travel and IT, people from our HRIS, CRM data, all of that comes into a single store, single
[19:53] source of truth on Databricks and AWS. We like the fact that the open architecture allowed us to seamlessly integrate from all of these different data sources stored in all of these different systems and bring that together so that the organization had that single consistent view across the board. And our teams can now build over
[20:11] dozens of applications that now sit on top of that same single store of truth to provide that unified view and applications across the board. So, that might mean that we have sales dashboards, it might mean that our people team is able to leverage our our to to manage our hybrid work policies.
[20:27] It might mean that our purchasing and procurement teams have greater visibility across the entire vendor set that they do. And all of those stakeholders can contribute their own data and build their own views and applications to suit their own needs. Any example? Glad you asked, Angie. So, uh here's a
[20:44] simple example. We have client briefs. So, everybody has the notion that your sales folks want to build better relationships with their teams. But, you have to do a lot of prep to make sure that you have visibility into the last conversations that you have, what their recent order patterns are, the kinds of things that are happening
[20:59] in the news and in the news cycle. So, we use our uh corporate data lakehouse built on top of uh, to bring together information from our CRM systems, billing opportunities, uh, product uh, launches, and real-time news updates to bring that together to provide our sales
[21:16] teams with that initial client brief that they can use to really focus on the relationship building rather than the mechanics of keeping that relationship alive. But, here's the one that our CFO is super excited about. We call it Beacon. Beacon is that sort of bright shining light through the fog, kind of like this
[21:32] light here that's shining on us through the San Francisco fog here. But, Beacon, if you look at it, you can see the menus across the top. What this does is it provides role-appropriate access to information that lets our executive teams, our C-suite, even our board members really look across the entire
[21:48] enterprise to get a sense of the business or to dive deeper into specific product lines and divisions for insights and trends. Our finance teams can run scenarios, they can look at performance of specific business units, and they can select filters to manage data visibility. If you If you took all of
[22:03] the information that's in Beacon and printed it out, it'd be 3,000 pages worth of information updated on a minute-by-minute basis, and our CEO and CFO look at this every single day. By having the data in one place, we have authoritative, governed, single source of truth for key metrics, models, and
[22:19] financial data that we use for our earnings reports, for conversations with analysts, you know, anytime that we publish data, it comes out through this through this uh, through this platform. And in fact, CFO Our CFO, Sarah, says if a KPI doesn't exist in Beacon, it really might
[22:35] as well not exist at all. I'm actually really proud of CFO. Usually, CFO only talk about money, not like platform investment. He actually saying, "You have to put things in Databricks cuz we have this platform." So, because she smells ROI.
[22:51] Um, so one thing that I just wanted to kind of call out is that every enterprise have gone through journeys of platform transformation and business transformation, acquiring the company, etc., etc. So, our core data comes in all different sizes, like structured,
[23:08] unstructured, on the cloud, or on the prim, and everything. So, Databricks really comes in really, really well in the heterogeneous environment to help us. Really two things that really you are calling out. Do you need a catalog? That's right. And which like enable us to not doing shadow copies, and the other thing is
[23:25] the open source and open open standard, which is really uh means no vendor lock-in. I actually we we got this script before like the CCC, the choice. Like it was like we are so much aligned with Databricks talking about.
[23:40] So, wonderful, Edwin. Well, no, they've been great partners for us. And you know, as we were saying earlier, as we transition into the AI phase, you really can't have a successful AI strategy without a successful data strategy. So, when we think about the AI that we're putting into our products, for example, that's built on that gold source data, that
[23:55] really unique asset that we have. For example, we launched last year uh Verafin's Agentyc Workforce. Verafin is our anti-fraud unit, and they take a rich view of consortium data by thousands of financial institutions that have opted in to to uh donate their
[24:11] transactional information into this pool, and we're using that to be able to look at fraud that spans the entirety of the financial ecosystem, not just an individual institution. We have agents that are on the lookout for new changes in regulation. They're not only helping us to understand that we may need to
[24:27] update our product, but also which of our specific customers are going to be impacted by that regulation change, and being able to bring that out to drive better relationships with our customers. And I know, Angie, you have a particularly relevant example from eInvestment, but I don't know that many people know what eInvestment is.
[24:43] Yeah, actually raise your hand besides folks from Nasdaq. If you know what eInvestment is. I see zero god, I expected zero. That those two deserve like, you know, great. So, but let me I think you probably should know investment. Here's
[24:58] like who has a retirement plan, a pension plan, or a endowment? Like probably 100% right? So, investment is the leading industry institution investment intelligent platform. So,
[25:14] it's like we hire whatever advisors to help us. So, institution folks, you know, for mutual funds or a pension fund, they actually hire asset managers, asset owners, and leverage investment platform to do the
[25:29] wise investment for all of your retirement funds or whatever. It has covered more than 90 trillion institution asset represented. 90 trillion. Think about what SpaceX What was the value of SpaceX? Just kind of
[25:45] compare to 25 years Like this is 90 trillion and we have more than 30,000 public a strategy and 90 90,000 or 80,000 private market funds. So, the important thing is that for asset manager asset owners in the world of AI,
[26:02] what they need to do? They actually want to use our investment platform, the data, and their own data in their own space to actually do AI analysis because making investment decision for a lot of money it is so incredibly important.
[26:19] The however part is that a lot of data aren't unstructured data, whether it's PDFs, charts, manager comments. And guess what? AI is really bad at it because there is no context. There is no the understanding what that data really
[26:35] for for all the connection part. So, our investment team, they created a patent pending method called Q&A method. They convert these unstructured data into context-aware
[26:51] AI-ready data. Now, with the AI-ready data putting in the client environment, particularly in the Databricks Databricks environment, magic happens. Yeah, that's right. So, you know, this is an example where we're taking our data, our gold source data, and clients
[27:07] are able to combine that with their own data sets in their own environment and start to run scenarios and strategies over that. But, it's not just data that just sort of lands in a data lake over there. It's actually because we're using Databricks and we're using the Delta Share technology, it actually now is able to interact with the entirety of
[27:24] the Databricks ecosystem that they're using. So, they can use Unity Catalog to have that level of governance and visibility and discovery over that data. They can use Agent Bricks and Genie to infer over that data and be able to talk to that and bring insights forward. They're able to use MCP servers that
[27:40] they own and the their inference model of their models of choice through the AI gateway layer to be able to bring their own intelligence across that data. And as a result, what we've found is that those folks are able to get to more accurate decisions, up to 50% more
[27:56] accurate, with up to 70% greater token efficiency because it's not just isolated data that's sitting there. It's a like Ali says, that context comes with that data and is able to then inform the LLMs and can come up with better, faster decisions. Yeah, the investment is providing the
[28:12] semantic layer really help on making decision faster and better. But, you know, I think, you know, there were only two people here I think that knew Investment, so maybe we could do a demo to help folks understand what we're talking about. Okay, let's go. All right, let me see if I can get this to work. Live live browser session here, so we'll see
[28:29] how here's the thing when Adnan is doing this is that when we have the AI-ready data, so there's there's there's a one things that we're actually trying to do. As an asset manager, we want to say, "What fund for for pension fund of actually what customer I'm going to invest?" And
[28:44] so, first you need to just making sure they're getting all the data to the data Databricks environment. What you're getting we have the in the private market space private market data strategy with performance and uh the the market lens for actually demand
[29:03] of of the pension fund. And the third one is the AI ready data. So, then you asked immediately when you have the data, you go to Databricks environment. We can pick like Genie, whatever. Now we actually pick the the playground. Um
[29:18] immediately what you can do with the data, like you ask a question, "Hey, you know, what what what are the customers that actually fit into in this in this category? And what is quality of the customer? And what is actually
[29:33] a buying opportunities? And because of AI ready data with actually connecting all the context, you're able to actually see the benchmark. So, that the asset manager can make decisions should I invest in this for the pension funds. And this is literally it can be done in
[29:50] like a few minutes after you get the data in. And again, this the credit I just want to pause a little bit. Investment team make AI ready data. And suddenly the intelligent come, the credit actually goes to the Databricks
[30:07] ecosystem. When you put the data over there, the Databricks ecosystem actually have AIs, data connectors, and everything to make things um easier for our client to use. As a matter of fact, our client is saying,
[30:22] "You guys really have a raised the bar for us to make our job easier and efficient." That's right. This looks really complicated, so we'll go back to the uh slides here and let folks know that if they want to hear hear more about investment and what you just saw there, they can come to our booth at the
[30:38] Marketplace Expo, booth number 157. Did I get that right? I think I got that right. Um where you can also hear about our data business and data link and their uh the recently launched MCP servers in private preview. Cool. All right. So. Um and today I mentioned that you just
[30:55] heard a slice of story. There's just so many AI story details and that just don't have time to to share. But that story is telling us is really a lot of things really start with the data. Start with like a strategy. And we start with solving a
[31:12] specific business problem into a data platform problem into the entire division and extending to really the entire enterprise for for Nasdaq. Now, this data strategy really now we've been getting the benefit of AI because of all this platform we're getting. So, what
[31:29] we're showing is that it really Databricks democratized the data and AI for our entire ecosystem. That's right. So, that's a little bit of a flavor of how a regulated financial institution like Nasdaq can leverage our experience and our store of data to be able to
[31:45] modernize how we run our business and to fuel innovation in financial markets. And before we wrap up and take some questions if we have time, uh we thought that we would share a few sort of keynotes from our journey along the way. So, first off, you know, everybody talks about AI. We're talking about AI a lot
[32:01] up here in the conference. But one of the most important aspects of a successful AI strategy is a viable data strategy. Whether it's for training or rag or context, the availability, quality, and and uh governance across that data is going to have a direct ability on your AI's ability to generate
[32:19] good answers. And being able to separate out the authoritative useful data from something that somebody put on a SharePoint somewhere or on a wiki, that's going to be really really key. That also helps you to identify where your data is most valuable. I talked about the uh Verafin consortium data set or our
[32:35] market data. These are key assets that we have that are irreplaceable anywhere else. And being able to understand what that is and leveraging that to drive your business value is particularly critical. As is building that semantic layer across the top because again, we talked
[32:50] about the data as the enabler, but the intelligence comes and brings that value. Well, like Ali said, this truly intelligent problem is actually context problem. But I wanted to really bring to you another story, another theme that we bring over here is
[33:06] that in order to scale this incredible technology, whether it's a data and AI, in an enterprise, you really need a strategy. The first one of the one of the principle we have the federated model, think about platform. If you have all the silos, you you create a lot of
[33:22] churns and inefficiencies. And the second, I'm going to skip a little bit, second, this is really not a technology. This is not a technology transformation only. This requires enterprise connections and the support from business and alignment.
[33:39] So, what we usually say that, you know, a lot of time is really technology transformation, it's really a business transformation in enterprise. So, really work on not just have the principle from technology, but really work through the enterprise to get this done. That's right. And that's actually a
[33:54] really good opportunity for us to say a huge thank you to all of our colleagues back at Nasdaq, the people in our corporate data lake house and finance teams, the people in capital access platforms, all of our support teams who, without their help, we couldn't be up here on stage talking about all this great work. We also want to acknowledge
[34:11] our friends and partners over at Databricks. They've been really fantastic for us. So, uh Sophia and uh Lauren and Ron and Hung and Dave, all of you have been fantastic partners and we couldn't have done this without you. So, thank you very much.
[34:27] Now, you want to wrap? Let's wrap. All right. So, we wanted to leave you with just this parting thought. You know, you hear Ali talk about the fact that businesses don't have a entire a intelligence problem, they have a context problem.
[34:42] Well, in admin and Angie said, we don't really have a technology problem, right? In other transform, we really, you know, have a culture principle organization of business thing that really make this to happen. So,
[34:59] you probably see the three names right there. I'm going to really talk about the first, the beginning, the question like what we have in common. What admin and Angie in common with Ali Ghodsi? We just want to be closer to him, right? He's such a visionary person.
[35:15] Um so, raise your hand who actually got the answer. What do we have in common? Right there. Hint, it's a visual trick. Okay. So, I will Okay, that person.
[35:35] Keep going. Letter A. Okay, so A I, that is correct. Very good. All right. You got a prize. You can come up afterwards. Ali Ghodsi. Raise your hand. I thought there's a few hands raising over there. Well, thank you. Thank you. What's your name?
[35:51] Bing. Bing. Okay. Bing. Thank you so much. Okay, so I just want to conclude one thing. When I we share this this kind of question to Ali, you know, Ali said, "Absolutely not, Angie.
[36:07] I have AGI in my name." All right. AGI is already here. There we go. Thank you so much for spending the time with us today. Please fill out your surveys.

Learn more about the Databricks Data and AI platform.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.