Snowflake Summit 2026 | Bala Kasiviswanathan, Snowflake & Caitlin Halferty & Laura Safdie, Thomson Reuters

Clips
More from Snowflake Summit 2026

Bala Kasiviswanathan

VP of Developer and AI Experiences

Snowflake

Caitlin Halferty

Head of Data & Analytics

Thomson Reuters

Laura Safdie

Head of Innovation Legal

Thomson Reuters

play_circle_outline Snowflake Unveils CoWork and CoCo; Thomson Reuters Partners on a Governed Snowflake Data Platform

play_circle_outline Fiduciary-Grade AI for Law: Authoritative Data, SME Vetting, Security, Transparent Outputs to Address Professional Liability

play_circle_outline Semantic layer/ontology for Customer 360 and finance semantic intelligence

play_circle_outline Data governance accelerating self-service BI and faster innovation.

play_circle_outline CoCounsel Evolving into an Agentic, Verifiable Legal AI: Tackling Hallucinations, Bias, and Validation

play_circle_outline Plan to scale semantic and agentic capabilities enterprise-wide

Info
Transcript

Bala Kasiviswanathan, Snowflake & Caitlin Halferty & Laura Safdie, Thomson Reuters

Bala Kasiviswanathan

VP of Developer and AI Experiences Snowflake

Caitlin Halferty

Head of Data & Analytics Thomson Reuters

Laura Safdie

Head of Innovation Legal Thomson Reuters

In this interview from Snowflake Summit 2026, Bala Kasiviswanathan, vice president of developer and AI experiences at Snowflake, joins Caitlin Halferty, head of data and analytics at Thomson Reuters, and Laura Safdie, head of legal innovation at Thomson Reuters, to talk with theCUBE's Dave Vellante about how a governed data foundation makes the leap from AI experimentation to enterprise-ready agentic intelligence possible. Kasiviswanathan anchors the discussion around two announcements: CoWork, which gives business users governed data access through natural l... Read more

explore Keep Exploring

How do Snowflake’s CoWork and CoCo offerings relate to your partnership with Thomson Reuters, and how does that partnership enable delivering real-world, governed AI? add

How are "fiduciary-grade" standards applied to AI used by legal professionals, and what are the legal implications? add

What does democratizing your data with CoWork/CoCo and a governed semantic data layer mean for the business? add

Will this lead to self-service business intelligence (self-serve BI), or would that be too dangerous? add

What is the technical and architectural journey of CoCounsel—from its pre-ChatGPT, RAG-based origins to its current agentic, integrated form within legal content and firm knowledge bases? add

What comes after the semantic layer—how will building an enterprise-wide ontology that enables confident inference and action develop? add

bolt Powered by CUBE AI

Bala Kasiviswanathan, Snowflake & Caitlin Halferty & Laura Safdie, Thomson Reuters

search

Dave Vellante

>> Hi everybody. Welcome back to San Francisco. We're here in Moscone South and you're watching theCUBE's coverage of Snowflake Summit 2026. This is our sixth year, actually seventh year if you include 2020 with COVID, covering Snowflake Summit. My name is Dave Vellante. I'm super excited to have three great guests here. Bala Kasiviswanathan this year is the Vice President of Developer and AI Experiences at Snowflake. And Laura Safdie is here. She's the head of innovation at legal for Thomson Reuters. And as I tweeted out earlier today, my business friend Caitlin Halferty is here. She's the head of data and analytics at Thomson Reuters. Folks, great to see you. Thanks so much for coming in theCUBE. Awesome to see you again, Caitlin.

Caitlin Halferty

>> Thanks. It's great to be here.

Dave Vellante

>> Wow. Do you believe we're here at snowflake Summit? 20,000 people.

Caitlin Halferty

>> Yes.

Dave Vellante

>> Amazing. The conference is jammed. It gets bigger every year, Bala.

Bala Kasiviswanathan

>> Yeah.

Dave Vellante

>> Okay. So CoWork and CoCo, formerly Snowflake Intelligence and of course your coding tool. We want to hear more about that, but set it up in the context of Thomson Reuters and your relationship with that company.

Bala Kasiviswanathan

>> Absolutely. First of all, very excited to be here. It's my first Snowflake Summit and it's great to share the stage with all of you. And today, it's been a great day. We announced both the Snowflake CoWork and CoCo. And the way we think about it is CoWork is for all the business users and CoCo is for builders. And we are all about making AI real for businesses. And that's where I think the partnership with Thomson Reuters comes into play. All these tools for AI are not real until they are on a governed data platform. And Thomson Reuters is a great proof of how they spent years building that trusted governed data foundation. Over 37,000 tables, 350 databases. And that foundation allows them to do some things amazing, and at a speed at which they want to innovate. But at the same time, if you look at how they develop and deliver the products, it's gated on very responsible AI measures. And that's what we love about this partnership, which is making things not about experiment or demos or pilots, but about how do we deliver innovation in the real world at a really significant pace.

Dave Vellante

>> Well, thank you for that, Bala. So Caitlin, you're up on stage today with Christian. I also tweeted out, "Get ready for the energy from the highly caffeinated Colombian, Christian Kleinerman." His energy is just amazing. You can't help but be excited when you're up there with them. But you had this concept of fiduciary grade.

Caitlin Halferty

>> Yep.

Dave Vellante

>> What does that mean?

Caitlin Halferty

>> Yeah, great question. So if you think about the professionals that we serve, they're lawyers, tax accountants, audit professionals, and they really carry professional liability for the decisions they make. And we want to deliver products and services that meet that highest bar of trust. And we really think fiduciary grade inclusive of four components. So one, it's the authoritative content. It's the investment in data security and safety. It's the subject matter experts that vet and validate the work. And then of course it's transparent outputs that we're able to validate. And I think the work that we're doing to build and deliver that in market is really what differentiates our approach and our execution results. I thought it might be interesting to hear from Laura on how we're using that capability in legal field as well.

Dave Vellante

>> How do you apply that? And I'd love to understand what the legal implications are there.

Laura Safdie

>> Yeah, they're enormous. So obviously AI has huge implications for the practice of law. But when you go to a lawyer, when a company goes to a lawyer, you're going to them for their substantive expertise, for their knowledge of the law and how to apply it to your situation. You can't work with AI that isn't rooted in verifiable legal data. So the difference between using AI that's unconnected to legal data, unconnected to legal expertise can mean getting it wrong. And so we work with the professionals who need to get it right every time. And so that's what it means in practice. Courts using AI, your lawyer using AI, it has to be rooted in authoritative legal knowledge.

Dave Vellante

>> Okay. So the legal industry is one that obviously is brought up often in terms of AI disrupting that industry. But to what you just said, Laura, I mean, I know when I talk to my legal professionals, I'm oftentimes looking, they know me, they know our business, and they oftentimes will give us advice that I won't get out of an LLM. A LLM might be able to tell me the law. It may or may not be correct. So I have to ask you, what do you see happening in the legal industry? It sounds like you're saying, yes, everybody's going to use AI, but Thomson Reuters is going to help you use the AI in a responsible and accurate way.

Laura Safdie

>> Yeah. One thing we work on when we're talking to law firms or courts or government agencies, it's AI is not a monolith. You have to be working in AI that's fit for purpose. And so if it's not rooted in legal knowledge, it's not fit for legal use. Now, what's important about the lawyers that you work with, it's their human lawyering. It's their knowledge of you and your needs and your business and your industry. And so an AI is not going to bring that human layer. And what we're able to do is see lawyers adopt AI for the more mechanic parts of practicing law, capturing all the data and the knowledge and millions of emails, and then leaving the lawyer to be the one that connects with you, that stands with you in court, that gives you the support and counsel that you need when you're facing what could be existential challenges.

Dave Vellante

>> Okay. So you guys are a data company, Caitlin.

Caitlin Halferty

>> Yes.

Dave Vellante

>> A lot of times, customers would say, "Well, we don't want to spend money on AI. We got to get our data house in order first."

Caitlin Halferty

>> Yeah.

Dave Vellante

>> Many have said, "It's taking too long. We're just going to go for it and have the AI help us clean up their data." I'm presuming, given that you've got a relationship with Snowflake, that your data estate was in pretty good shape pre the AI heard around the world. How would you describe the way in which the Thomson Reuters brought AI? I know you haven't been there since 2022, but nonetheless, you've got some historical perspective on this that you can draw on. How would you describe the way Thomson Reuters brought AI into that data estate and how has it evolved since then?

Caitlin Halferty

>> Yeah. So I joined TR a year ago. And fortunately, my predecessors in that role really did the foundational work. The governance, figuring out our data domains, our critical data elements, bringing them into Snowflake, getting them governed, et cetera. When I joined, I really focused on a couple of things, making sure that we could accommodate for unstructured data. So all the work we did around structured and unstructured with the right row in column level security, really meaning we're giving the right access to people across the firm. And then we have a big initiative with our finance colleagues around semantic intelligence. So we've built this semantic capability on top of our platform, and really what that means is a business intelligence across all of our disparate data sources. So for example, we historically had about 23 different data sources we needed to pull from for customer master to understand our customers. We were able to bring all of that together. The other piece that we've really been going after is 2026 is our year of customer obsession. So really everything we can do to better understand our customer needs, be proactive, reach out. So again, leveraging the platform and the work we did to bring our data together. And this particular piece of work is around retention. So product usage, customer sentiment, billings collections, renewals, all those base signals are pulled together. And now as we run our alerting campaigns and we increase our digital footprint, we're able to enable our sellers to act effectively with that insight all pulled together. So that's been a couple of examples that really have driven significant business benefit for us internally.

Laura Safdie

>> And can I just say that as our users are navigating their own transformation of their professions, the fact that we have higher fidelity data on which we can assess how are they using our products, where are they in the learning curve, has totally unlocked a different way of communicating with our users to help them through their AI journey.

Dave Vellante

>> And that's what you care about as a JT.

Laura Safdie

>> That is what I care about.

Dave Vellante

>> Yeah, okay. I apologize if this is not true, but you're not a techie or are you?

Laura Safdie

>> I am not a techie, but I did co-found a tech company with people who were techies.

Dave Vellante

>> Okay, cool. So the reason I ask that is because, Bala, we use terms, like Caitlin just used semantics. We talk about catalogs, we talk about ontology. And a lot of times to business people, those don't mean anything. They care about data fidelity as an example. So tie that in, because we throw these terms around a lot. They're critical to you in terms of your product. How do you translate those into business terms and business value?

Bala Kasiviswanathan

>> Yeah. I think the good example here is that I think once you have the foundation there, I think what Thomson Reuters is trying to do is how do you open it up to every user to use all the knowledge they have?

Dave Vellante

>> Democratize it.

Bala Kasiviswanathan

>> Yes. And a great example, Caitlin was talking about it, is if you're running a Customer 360 and people have to understand what to do with their customers, and there the semantics means when I say revenue, do you mean revenue in the sense of every revenue professional out there in Thomson Reuters? That's what it means. And I think with CoWork and what you do is you take away all that complexity and let them talk in natural language and get what they need out of that, and somebody behind the scenes is mapping all those things so you know that you're looking at the right business term, you're looking at the right data, and you're also providing the right data to the right user at the right time. All of that is what is available through something like CoWork and it's possible because of what they've done with the data estate to make that happen. And in addition to that, they have over 1500 users using CoCo, and that allows them to also build new things on top of that data. And again, those people are builders and sometimes they are in a particular domain or a particular area of expertise within Thomson Reuters, but the governed data layer with all that semantic views and ontology allows them to build what they need to build with CoCo without having to worry about what exactly is each of these kind of data points referring to.

Dave Vellante

>> So what does that mean for the business, Caitlin? I presume it's less data wrangling, less arguing about what revenue means. All the coordination that goes into, you know the old Duke pipeline mess. And so it gives you the ability to spend more time on some of the more strategic and judgment level, but I wonder if you can-

Caitlin Halferty

>> Absolutely. And we intentionally partnered with our finance colleagues when we did this work. Because if finance says, "This is game changing, I trust these outputs, I'll sign off and we'll run our business and financial reporting externally off of this data," then you know it's at the level of accuracy that you need. And back to the fiduciary grade point from earlier, we want to model our same focus on really that high quality, right access data internally and model that with our customers.

Dave Vellante

>> And does it lead to like self-service BI or is that too dangerous?

Caitlin Halferty

>> No. I mean, we are very clear, within AI ... We've been early adopters of AI technology. We have cross firm initiatives in which everyone is encouraged, invited to share their AI output, what they've been able to drive in terms of business results and impact. And the work we're doing on the data estate is what is enabling that. And now as we introduce, for example, more agents, it's a goldmine. So we have curated appropriately governed data that agents can run on and instead of having wrong answers that come from some confusion or disconnect between data terms, we've done that hard work. So that's an important part of what we've been focused on. And we've been able to move very quickly in the last quarter.

Bala Kasiviswanathan

>> And self-serve is the future, and once you have that kind of secure governed platform, you can actually have people go and do what they need to do and remove all sorts of bottlenecks so that you can move fast.

Dave Vellante

>> Yeah. Governance is an accelerant. That's a new thing. Okay. But Laura, so we talk about CoCo. You have a CoCo too. Not quite, but CoCounsel.

Laura Safdie

>> CoCounsel, yes.

Dave Vellante

>> I think is what you call it.

Laura Safdie

>> Some call it CoCo.

Dave Vellante

>> I would, but I guess I won't now. But what is CoCounsel? How does it all fit in here?

Laura Safdie

>> So before the world knew about ChatGPT, we were building the world's first legal AI assistant-

Dave Vellante

>> Really?...

Laura Safdie

>> on top of that technology, yes. And since then it's grown in Thomson Reuters building agentically natively into the legal data, the legal expertise, and the data within litigation within a firm's knowledge base. And so what it's become is essentially a companion that follows a lawyer through every step in the legal journey, whether it's through intake, whether it's through briefing cases, through analyzing contracts, all of that can be enabled. And it allows lawyers to spend their time doing that human work. And one other thing I'll say is that kind of a personal passion is what is also unlocked is an incredible opportunity in expanding access to justice, seeing legal aid organizations and pro bono teams leverage AI to serve more people in need.

Dave Vellante

>> So, Caitlin, I think about our AI tools. I go back to like 2015 when we ... So you maybe remember when we started-

Caitlin Halferty

>> Automated metadata generation.

Dave Vellante

>> Yes. But I look back and, wow, that was pretty rudimentary compared to where we are today.

Caitlin Halferty

>> Yeah.

Dave Vellante

>> How would you describe ... So the pre-ChatGPT CoCounsel, what is that like technically and how architecturally did it evolve? Probably went through being a RAG-based chatbot and then you're injecting a lot more intelligence into it and maybe agentic. How should we think about that? Could you guys fill us in on that? Do you have enough technical detail? Or maybe Bala could help us.

Caitlin Halferty

>> Yeah. I mean, I think what you're asking, and you were talking about several years ago when we were developing those first AI solutions. We were doing-

Dave Vellante

>> Yeah, what's the journey that CoCounsel went through to get to where it is today?

Laura Safdie

>> I can talk about that.

Caitlin Halferty

>> Yeah, maybe I'll let Laura talk about CoCounsel specifically.

Dave Vellante

>> Okay.

Laura Safdie

>> So initially when we all saw ChatGPT-4, that was like a real inflection moment where we understood that we can point this incredibly powerful model at an authoritative source of data. So yes, RAG was an incredibly important initial tactic, but it was also the application layer which built in legal prompting that was at a level of expertise and domain specificity that unlocked uniquely legal work. Now we're in a different stage. We're building out and in the process of launching to market the next generation where we've basically rebuilt the brain within CoCounsel to be into an agentic harness that's natively built into the legal content in Westlaw Practical Law, which are the authoritative sources of content for the legal profession, as well as the knowledge at a law firm or in house legal team. And so building it agentically unlocks really complex work product in a way that earlier generations couldn't satisfy.

Caitlin Halferty

>> And Dave, if I could build on that, one way our teams intersect in that work, which I think is so important is just as Laura described, there's been this really incredible journey where we've pushed at the forefront of the capabilities of the product. We have a team, a responsible AI team, and before any sort of AI capability integrates the product and ships to market, we do all the robust analysis around hallucination, bias detection, anything that you're worried and concerned about in the umbrella of responsible AI. And so as it hits market, we have confidence that we have done that work and that analysis. And now we have such rich history of that data across our products as they've evolved, we've been able to really do some fantastic analysis of that as we look forward to these future releases that are coming, particularly our agentic release.

Dave Vellante

>> And Caitlin, I want to come back to that and I'll tell you where I'm going. I want to understand what your vision is beyond the sort of semantic layer that you've built. But I want to come back to Laura because you are actually kind of a techie. When you use the term harness, okay, what do you mean by that harness? When I think of harness, Bala, I think about like a coding copilot or a coding agent, but is that what you mean within CoCounsel?

Laura Safdie

>> Yeah. We built the next generation of CoCounsel on top of the Cloud SDK and it's put us in a position to leverage the authoritative data that Thomson Reuters has been building for decades, maybe almost over 100 years, natively into the agent, into the assistant. And so that way you can ask CoCounsel not just review these contracts, you can ask them develop a plan for this entire litigation and leverage the knowledge that's in various sources of data, both Thomson Reuters as well as the client as well as the legal matter. And what it unlocks then is an agent that can build a plan for you, but at every step of the way informed by legal expertise and also critically for fiduciary grade AI is verifiable at every step. So me, as the lawyer, I'm still in charge of my matter and I'm still reviewing every bit of that plan.

Dave Vellante

>> So go please.

Bala Kasiviswanathan

>> No, I was going to say, I think that the harness part is all about system prompts, rules, domain skills. That's what makes, I think, CoCounsel special. And in terms, you asked about where the future could be, the good news is the way I see Caitlin and Laura working together, we facilitated that even more with a couple of announcements we made today specifically. One is the Cortex Sense runtime and then the Cortex CoCo Agent SDK. So in the future, actually CoCounsel could actually use all of that data, all of that governed data and everything that you have in terms of context using an SDK into CoCounsel. So that's kind of the future and the out of the possible should you want to go there in bringing that domain knowledge even more.

Dave Vellante

>> Well, since runtime starts to get to my follow up to Caitlin, which is, okay, what's beyond the semantic layer? You got taxonomies, you got your catalog, you got the semantic layer. Now, we want to create this ontology and we want to infer and act confidently. So where do you see that going?

Caitlin Halferty

>> Yeah. So we've delivered in production semantic for primarily our finance community and organization. We want to quickly move to enterprise scale and so that's our immediate action, and I think we've seen incredible results. And now that we trust the output and we can vet and validate and trust that, we want to move to enterprise scale. And then that's going to be another way in which we accelerate our AI agent work. So we've been looking a lot at what does a human agent interaction need to look like, and some of the announcements today help us with that as well so that we can effectively come with that equation and quantify the impact of that piece of work. So that's the next piece, expand to enterprise and do the agent piece.

Dave Vellante

>> So Reuters was founded in 1851, and Thomson Corporation, 1934, and Thomson Reuters April, 2008. So you're not a startup. So you don't have a greenfield blank sheet of paper to get AI-afied. What I'm interested in, so I see fiduciary grade is kind of your unique value proposition. That's what you do better than anybody else and you're like laser focused on that. I'm curious as to how an organization like Thomson Reuters becomes so called AI native. You can't just say overnight, "Okay, we're going to be AI native." It's too large. So my thinking is you go at it with edge cases, maybe like legal, but you start with a data foundation. Is that how you see ... Well, first of all, do you want to be an AI native company? It's kind of a silly question. I think everybody is thinking that way. And then how do you get there with a company that is as established as Thomson Reuters?

Caitlin Halferty

>> Well, I'll let Laura weigh on her perspective, but I think the acquisition strategy we have is an important one. We are very acquisitive as a firm and we like to acquire really deep, important differentiated tech and talent such as Casetext's acquisition that Laura joined for. So I think there's an acquisition component to it as well. And yes, we have rich legacy. And with that, we have customers that have been loyal with us to our products historically, which is amazing and we're in an increasingly competitive environment. So always asking ourselves, what can we do to deliver best product for today? And I think that there's a M&A component to it. There's an investment in innovation that we contend to prioritize as well. But curious what you might think.

Laura Safdie

>> Right. So I co-founded Casetext before it was cool to do professional AI, and I will say that the reason that we merge with Thomson Reuters is the recognition that you can bring the best AI talent to bear on a problem, but if you don't partner that with authoritative, verifiable data, you're just limited in the impact you can have for individuals and industries. And so that's why we joined Thomson Reuters and gives us an opportunity to do things on a scale and at a level of impact that we never could have imagined doing on our own. And like you said, Thomson Reuters is acquisitive. We brought on Additive and Pagero and Noetica, all very innovative companies giving professionals service that's rooted in data. And I will say, I think that our cultures and our leadership teams have come together with a new wave of leadership at Thomson Reuters that's all about moving fast and aggressively with AI, but also never losing sight of our north star, which is fiduciary grade and trust for the professions that we serve.

Dave Vellante

>> So that's amazing. Congratulations-

Laura Safdie

>> Thank you....

Dave Vellante

>> on that. And that's what I sort of consider. So it's almost like, maybe it's not the best example, but when IBM bought Red Hat, it was like, we're going to inject the Red Hat-ness into IBM and that opened, right? I mean, as opposed to the old stodgy IBM. But by bringing in startups that are AI savvy, those I see as sort of edge cases that the core can learn from and that knowledge or that discipline sort of migrates over. You can't do it all at once. So it seems to me, Caitlin, that your challenge, as the head of data, is to make sure that I don't just create silos of automation, that it has to be an organizational substrate that everybody can take advantage and that doesn't happen overnight. How do you see that journey?

Caitlin Halferty

>> And you're exactly right. Especially as we continue to acquire the companies that we just talked about, there has to be an integration path. We want to leverage the customer relationships that those firms bring us and make sure we don't disrupt that service. And what's actually put us in a good position to do is the investment that we have in the data foundation and the ability to integrate and connect and build. And then we're doing our big commercial systems design and delivery as well and really examining all of the tools that we've integrated. And some of these we can build ourselves now, which is great. So we've been able to build on top of and I think I would say we have a culture of people that gravitate toward tech. We have builders, they're eager to contribute in that way. We have some really good examples of building those capabilities.

Bala Kasiviswanathan

>> So I like to pair what Caitlin said, I think which Sridhar said yesterday very well, which is for a company like Thomson Reuters, I think your customer data, your operational data, your knowledge, that becomes the moat and that is a big one.

Caitlin Halferty

>> Yeah.

Bala Kasiviswanathan

>> And he was mentioning this and I can see that as a great example of how Thomson Reuters puts it to practice. So you are to bring AI to that data, which is your unique moat, unique differentiator. And I think the other thing, what do you see here is I think Caitlin and Laura actually also showing that innovation is a mindset and I think by opening up access to tools and allowing people to go and experiment, and having the right gates to have responsible AI, I think that's why AI means the data and that's what innovation looks like.

Dave Vellante

>> I do think it's all about the data and I'll leave you with something else, Sridhar said last night to us, because people are very confused right now. You look at the market, oil's up, oils down, AI, ROI. When you go through these transitions people, they get confused, I'll just say that. And Sridhar said, "It's really hard to predict who's going to win, what's going to happen." Whoever wins, they're going to rewrite history and say it was always going to be this way and they'll claim victory. That's how this industry works." But in an homage to Frank Slootman, he said, "Product market fit is a life force."

Bala Kasiviswanathan

>> Yes.

Dave Vellante

>> And I won't forget that because it's true. If you're confused, just build great products, stay close to your customers, and you win. Caitlin, give you the last word. I'll go around the horn for the last word, you guys.

Caitlin Halferty

>> Sure. I think it's been incredible what we've been able to build in a very short amount of time, and delivering the semantic work has been a great milestone. I want to move that to enterprise scale and we have a plan for that. The work we're doing to really build the curated gold mine for our agents and the focus on data I think will serve our internal transformation story, as well as the products that we ship to market. So it's really, I think, been an example of what we've been able to do very quickly, both internal and external.

Dave Vellante

>> And Laura, I see you as sort of the green shoots coming into Thomson Reuters to bring that innovation and bring that sort of AI native culture and mindset. What are your final thoughts?

Laura Safdie

>> I mean, there are partners and leaders across the organization of the same point of view and we're partnering across the organization. I will say you made a point that data is king, but context is queen. And it's not just the data that we have, it's really the legal expertise. It's understanding what it means to practice law, to serve a client, to work in the interest of justice. And you can't just bring AI to an industry you don't know, you can't just transform a workflow you don't understand. If you are building for one of the most important professions in the world, you need to understand how AI can make us better and serve the public interest, while also doing it in such a way that's fit for purpose. And that's kind of what drives us every day.

Dave Vellante

>> And Bala, that is a great point that Laura just made and that's really where things like Sense come in to get that context to understand the workflow and the process knowledge. I mean, three years ago, four years ago, that's not something that Snowflake was even thinking about, and now you're clearly moving into new areas and that sort of is the future of the company. I'll give you the final word.

Bala Kasiviswanathan

>> Yeah. No, I think, look, at the end of the day, data, like in a great partner like Thomson Reuters, is their moat, but once you have governed data, which you trusted and then you have the context built around that with something like Cortex Sense, I think the opportunity is huge in terms of what we can do together. And I'm very excited about, with the tools like CoWork and CoCo, what Thomson Reuters can unleash within the organization and for their customers.

Dave Vellante

>> Guys, thank you so much. Very educational and inspiring for our audience. I hope it helps them take some action and appreciate your time. And thanks for coming in theCUBE.

Laura Safdie

>> Thank you.

Bala Kasiviswanathan

>> Thank you.

Caitlin Halferty

>> Thank you.

Dave Vellante

>> All right. Thank you for watching. This is Dave Vellante, for Rebecca Knight. We're here live at Snowflake Summit 2026 from Moscone. We'll be right back for day one coverage right after this break.