Snowflake Summit 2026 | Artin Avanes, Snowflake & Nancy Wang, 1Password

Clips
More from Snowflake Summit 2026

Artin Avanes

Head of Core Data Platform

Snowflake

Nancy Wang

Chief Technology Officer

1Password

play_circle_outline 1Password as the Internet's Vault: Governing Humans, Machines, and Agents Beyond System Security

play_circle_outline Layered Security for Agents: Semantic Layers and Horizon Context to Boost Accuracy, Trust, and Data Protection

play_circle_outline Agent identity concept introduced by Snowflake

play_circle_outline Collaboration to close security seams and build end-to-end lifecycle integration

Info

Artin Avanes, Snowflake & Nancy Wang, 1Password

Artin Avanes

Head of Core Data Platform Snowflake

Nancy Wang

Chief Technology Officer 1Password

search

Rebecca Knight

>> Hello, everyone. And welcome back to theCUBE's live coverage of the Snowflake Summit 2026 here at the Moscone Center in San Francisco. I'm your host, Rebecca Knight, alongside Dave Vellante, co-founder, co-CEO of theCUBE. I would like to welcome two CUBE veterans to the show. Artin Avanes, head of Core Data Platform at Snowflake, welcome back.

Artin Avanes

>> Always great to be here.

Rebecca Knight

>> And Nancy Wang, chief technology officer at 1Password.

>> Thanks for having me back.

Rebecca Knight

>> So thrilled. For our viewers who are less familiar with 1Password, tell us what you're all about and what you do.

>> Yeah, sure. So most folks know us as a consumer platform, and I'm here to say that we actually have a really big B2B business as well. And so we like to call ourselves the vault for the internet, which means that today we store more than 1.3 billion credentials across humans, machines, think service accounts, and now, of course, AI agents. And so we're the source of identities, and also we broker trusted access.

Rebecca Knight

>> So how would you say the security conversation has evolved among enterprises and in the B2B space, for example?

>> Yeah. So I feel like for many decades, like the last 20 years, it's all been about how do we, for example, secure applications, cloud infrastructure, on-prem infrastructure. And the talk, especially at the summit this week, and I'd love to hear your thoughts, has been around, well, how do we govern actors who are trying to get access to those sensitive systems, that sensitive data? So the conversation has really changed from how do we secure systems to now how do we govern actors. And I'm sure we're probably going to get into it, but agents really throw that conversation and make it truly multidimensional.

Dave Vellante

>> So there's definitely a workload shift going on. As Nancy just said, it's kind of agentic with Snowflake. I mean, at some point, it's probably going to become increasingly accelerated, but maybe that's a little bit down the road. You have so much work to do now. But how has that affected the way, Artin, that you have to think about security?

Artin Avanes

>> I mean, Nancy was just talking about it, or alluding to it. In a world where you have to expect thousands of agents accessing your most important assets, your enterprise data, we have to deal now with a very different order of magnitude. Thousands of agents are accessing your data. How do you really, A, keep tabs on what agents exist in your system? So the observability part, the transparency part. And then how do you actually apply the right governance and security controls? There is no world where you can just put an agent out or run your agentic tool directly against your data without those controls. It would destroy your business in many ways. So that is really what we have been hearing from our customers and why we have responded the way we have responded in this summit with a slew of new capabilities that allow you to control and govern the access.

Dave Vellante

>> It was interesting. We were just talking to Christian about the emphasis that you're putting on things like business resilience. And the definition of that resilience, cyber resilience has really, really changed. What role do you play in that conversation? Because we kind of came out of COVID saying, "Oh, wow. We're not that resilient," and then sort of caught up a little bit. And then the GenAI hits, and it creates a whole new equation. How do you think about that?

>> Yeah. Well, look, resilience, I feel like that was the topic for our last conversation, right?

Dave Vellante

>> Yeah.

>> And so prior to 1Password, I was a general manager for data security and ran data security products at AWS. And the conversations we had back then were all around, well, cyber resilience. What do you do, especially after an attacker gets access to your sensitive data? Well, I think to our conversation, maybe especially so this week during summit, it's now evolved into, how do you prevent unauthorized access, right? And so you actually block attempts that are not intended for sensitive data before it even happens. And so that goes back to, how do you think about cyber resilience, going back to defense and depth? And it really starts with identity now as a security perimeter and identity as a control plane, which is who the actor is, right? Who authorized that actor? And then, of course, now going to the data side, which is, what is that actor now allowed to do with that sensitive data, whether that actor's a human or an AI agent?

Dave Vellante

>> So what I've heard from a technologist is you got to protect the agent. That's where the focus should be. I've heard, "No, no. You got to protect the data.2 What you're saying is both. It's got to be-

>> Yeah....

Dave Vellante

>> layered approach.

>> And that's actually one of the reasons we're here today is talking about, for example, how our two companies can come together around securing the agent identity, and then, of course, with Snowflake, securing and governing the data access.

Rebecca Knight

>> So, Artin, you were talking a little bit about the different kinds of requirements that customers have, things that they're asking for that they maybe weren't a year ago. How do you describe the ways in which AI workloads have fundamentally changed expectations around performance and ease of use and security?

Artin Avanes

>> I think it starts with accuracy. So really, if you look at our customers who have deployed now agents at scale, it comes always down to how accurate and trustworthy those agents are, right? And there are different dimensions to it. I would say the capabilities we offer through the semantic layer that we introduce are a critical building block to make really AI successful at scale. If you don't have a shared understanding of your data, it's slightly autogonal to security itself, but it is a critical piece when we talk about productivity and accuracy of these agents. The semantic layer is a critical piece to it. And we announced at summit additional capabilities, like Horizon Context, that really taps into metadata that sits outside of Snowflake to also, again, power your agentic workflows and processes much more accurately moving forward. So I think that is really an important pre-req for our customers to actually deploy agents and AI applications at scale with success.

Dave Vellante

>> Do you tap into that, Nancy? Like, what parts of the Snowflake stack matter to 1Password? What are the most salient aspects of it that-

>> Yeah....

Dave Vellante

>> have value?

>> So this goes back to our ethos, which is as the vault for the internet, we contain sensitive credentials, things like API keys, tokens, SSH keys, so on, so forth. And so from a data warehousing perspective, Snowflake has incredible amounts of super sensitive and top-secret data for many enterprises out there, including 1Password. And for many of our customers who also use Snowflake, the question becomes, well, who is that agent? Who delegated authority to that agent to be able to access Snowflake to, let's say, create an analytics dashboard or be able to answer important business questions? And so that goes back to the originating identity of that agent. And so there are ways that, for example, we can bind intent cryptographically to that agent identity such that as that identity is issued to the reasoning layer, to now the data access layer, that intent stays pristine. And you don't have drift from the original intent for which why this agent was created in the first place. And so that's the work that I would say alluding to what we're doing together, and it's so important. Because going back to your earlier question, it's really a case where identity security now meets data security, and you have to preserve that intent across each layer of the stack.

Artin Avanes

>> I would say accuracy was one piece where the semantic layer comes in the picture. Now, with respect to how we integrate with partners like 1Password, it's really around our start of the notion of identity. I mean, everyone is in the industry trying to figure out the story around agent identity, and we introduce at summit the notion of an agent identity that helps our users to differentiate between a human user and a actual agent user. So that's where we can then work very closely with the broader security ecosystem to really secure agents and what they can do. Because the notion of an agent identity is just a building block. With an agent identity, then you can combine it with access policies or security policies that allow you to protect your sensitive data and allows you to define what an agent can do or cannot do relative to a human.

Dave Vellante

>> So agents are onboarded and offboarded a lot faster than humans are. So how do you think about handling that dynamic?

Artin Avanes

>> It comes down to a couple things. First of all, I would say there's still a friction, specifically when you think about highly regulated customers on onboarding of agents. So it really depends on the customer vertical that you're talking about. There's still a lot of friction when it comes to, can I deploy hundreds or even thousands of agents seamlessly? And we are trying with our new AI security capabilities to accelerate that emotion. So really about how can you build that trust to actually allow customers to deploy at greater scale. Now let's talk about the customers that actually have already deployed and might have gone through that motion fairly quickly. They come back and say, "Well, I want now to better understand what's going on. Who is doing what?" So that's why we also introduced our extension to Trust Center, which is our security product within Snowflake, which we announced in Christian's keynote, which is around the security posture for all of your agents that gives you visibility of what agents are running, under what role, under what permissions, what scope they have. Are there potential violations or warnings that we should surface to you automatically? So these are, I would say, the different spectrums from highly regulated ones where the onboarding-offboarding is actually not easy towards the ones who actually have onboarded very quickly, but then come back and say, "Wait, I want to actually manage and govern that better then."

Dave Vellante

>> But really different paradigm than human-

Artin Avanes

>> For sure....

Dave Vellante

>> security versus...

Artin Avanes

>> For sure. And that's why I'm saying the ones that are highly regulated are concerned about, well, now you can easily deploy hundreds or thousands of agents. It's not anymore as slower with the humans, but that's where the friction comes in the picture. And that's what we are trying now to help accelerate with our new AI security capable.

Dave Vellante

>> What are you telling customers about PQC, post-quantum cryptography? Is it something they should be paying attention to?

Artin Avanes

>> I would say yes. The question is always the time horizon. Is it something that would be extremely relevant short term? I would say probably not. Versus long term, definitely yes. So it's really about more around the time horizon than-

Dave Vellante

>> And I think it also-

Artin Avanes

>> Oh, go ahead....

>> relates to where that cryptography is put into place. So for example, as 1Password, we take our mission for protecting credentials very seriously. And so that's where maybe we'll come up with a recent white paper on how we're going to future-proof our vaults or vault encryption against future quantum.

Dave Vellante

>> Because people, hackers, will exfiltrate cryptographic or data that's encrypted with the hopes that someday they'll be able to hack it with quantum, even though it might not be here for five years or however it is. So in your white paper, you talk about how you're essentially solving for that now with the expectation that at some point it's going to be exfiltrated. Is that right?

Dave Vellante

>> Future-proofing, because... you know, and won't name names, but certainly in the industry, right? There have been cases where vaults have been, let's say, brute-force spammed and sensitive credentials get leaked. And so that's one thing that we're super proud of internally, which is we have a zero-knowledge architecture that we design or put all of our products through. And what that means is, even if we were to get breached, an attacker would never get access to sensitive credentials that are stored within our vault.

Dave Vellante

>> Yeah. So you can't even get to them, right?

Dave Vellante

>> We can't even see them. That's actually what we say, right? Attacker can't see them because we can't even see them.

Dave Vellante

>> It's a very AWS mindset from your days there.

Dave Vellante

>> Exactly. We actually built it on AWS Nitro. We're a poster child for-

Dave Vellante

>> Yeah. Okay....

>> Nitro Enclaves.

Dave Vellante

>> Yeah, that makes sense.

Rebecca Knight

>> So, Nancy, what is standing out to you at this summit in terms of announcements that Snowflake has made at a time when AI adoption is increasing so rapidly? What is standing out to you about its approach and its strategy here?

Artin Avanes

>> Yeah. So maybe going back to what Artin said around announcements such as agent observability, different security controls. I think the theme for me that really stood out this week at summit is around the agentic enterprise, right? The conversation has really moved beyond, "Hey, let's secure models." That was really, I think, gen one of AI security. And now it's more agents are here. They're in production, right? We're talking about agent swarms, hundreds, thousands of agents. In fact, I just spoke with agent sandbox company that says they can now spin up concurrently one million-plus different sandboxes. I mean, just think about that scale. And so when you have that scale... And specifically for this customer, they're building this agent swarm to do DevOps work. So if an agent did an update to a database or maybe did a config change, how can you actually map that trust and accountability back to that specific agent in a swarm of millions of agents? And so this goes back to what Artin just said around security controls and observability tying back to accountability.

Dave Vellante

>> Where do you want this relationship to go? What should we expect a year from now we're back at Snowflake Summit 2027. What do you guys want to be able to say that you can't say today?

Artin Avanes

>> I think I would love to be able to cover the entire life cycle from a security perspective. As Nancy was saying, it starts always with the identity and goes then down to the data access. So the entire end-to-end life cycle and building a seamless, very secure integration. As we are bringing the new Snowflake native capabilities, we will always need to have the broader security ecosystem integrate well. Because we can introduce a bunch of native security capabilities. If the ecosystem is not well integrated, there's always risk, right? And I think that is, high level, I would say, the North Star that we want to work with and get also feedback from our partners to say, "Look, this is areas that we need to jointly invest." It's really a joint effort, and it's not one tech partner or vendor solving itself.

Dave Vellante

>> So closing those seams with collaboration and presumably standards that-

Artin Avanes

>> Yes. Yes....

Dave Vellante

>> help you close those gaps.

Artin Avanes

>> And again, the entire life cycle and workflow when it comes to the propagation of the identity down to the data systems like Snowflake.

>> Yeah. I mean, very simple. Maybe one-liner, which is 1Password and Snowflake helps keep your data secure and secures your agent access to data. And this goes back to, I think, Snowflake, you're one of the most innovative customers that we have, which is... I heard from Sridhar, actually, during HumanX conference that you've already built skills using our 1Password developer tools.

Dave Vellante

>> Yeah.

>> And so I think that forward-thinking around the agentic enterprise and everything that you've launched at summit carries through some of the security work that we're about to do together.

Rebecca Knight

>> Excellent. Well, Nancy and Artin, thank you both so much for the conversation. Really interesting.

Dave Vellante

>> Good to have you. Thanks.

Artin Avanes

>> Thank you. Thank you for having us.

Dave Vellante

>> Thanks for having us.

Dave Vellante

>> You bet.

Rebecca Knight

>> I'm Rebecca Knight for Dave Vellante. Stay tuned for more of theCUBE's live coverage of the Snowflake Summit. You're watching theCUBE, the leader in enterprise tech news and analysis.

Artin Avanes, Snowflake & Nancy Wang, 1Password

search

Rebecca Knight

>> Hello, everyone. And welcome back to theCUBE's live coverage of the Snowflake Summit 2026 here at the Moscone Center in San Francisco. I'm your host, Rebecca Knight, alongside Dave Vellante, co-founder, co-CEO of theCUBE. I would like to welcome two CUBE veterans to the show. Artin Avanes, head of Core Data Platform at Snowflake, welcome back.

Artin Avanes

>> Always great to be here.

Rebecca Knight

>> And Nancy Wang, chief technology officer at 1Password.

>> Thanks for having me back.

Rebecca Knight

>> So thrilled. For our viewers who are less familiar with 1Password, tell us what you're all about and what you do.

>> Yeah, sure. So most folks know us as a consumer platform, and I'm here to say that we actually have a really big B2B business as well. And so we like to call ourselves the vault for the internet, which means that today we store more than 1.3 billion credentials across humans, machines, think service accounts, and now, of course, AI agents. And so we're the source of identities, and also we broker trusted access.

Rebecca Knight

>> So how would you say the security conversation has evolved among enterprises and in the B2B space, for example?

>> Yeah. So I feel like for many decades, like the last 20 years, it's all been about how do we, for example, secure applications, cloud infrastructure, on-prem infrastructure. And the talk, especially at the summit this week, and I'd love to hear your thoughts, has been around, well, how do we govern actors who are trying to get access to those sensitive systems, that sensitive data? So the conversation has really changed from how do we secure systems to now how do we govern actors. And I'm sure we're probably going to get into it, but agents really throw that conversation and make it truly multidimensional.

Dave Vellante

>> So there's definitely a workload shift going on. As Nancy just said, it's kind of agentic with Snowflake. I mean, at some point, it's probably going to become increasingly accelerated, but maybe that's a little bit down the road. You have so much work to do now. But how has that affected the way, Artin, that you have to think about security?

Artin Avanes

>> I mean, Nancy was just talking about it, or alluding to it. In a world where you have to expect thousands of agents accessing your most important assets, your enterprise data, we have to deal now with a very different order of magnitude. Thousands of agents are accessing your data. How do you really, A, keep tabs on what agents exist in your system? So the observability part, the transparency part. And then how do you actually apply the right governance and security controls? There is no world where you can just put an agent out or run your agentic tool directly against your data without those controls. It would destroy your business in many ways. So that is really what we have been hearing from our customers and why we have responded the way we have responded in this summit with a slew of new capabilities that allow you to control and govern the access.

Dave Vellante

>> It was interesting. We were just talking to Christian about the emphasis that you're putting on things like business resilience. And the definition of that resilience, cyber resilience has really, really changed. What role do you play in that conversation? Because we kind of came out of COVID saying, "Oh, wow. We're not that resilient," and then sort of caught up a little bit. And then the GenAI hits, and it creates a whole new equation. How do you think about that?

>> Yeah. Well, look, resilience, I feel like that was the topic for our last conversation, right?

Dave Vellante

>> Yeah.

>> And so prior to 1Password, I was a general manager for data security and ran data security products at AWS. And the conversations we had back then were all around, well, cyber resilience. What do you do, especially after an attacker gets access to your sensitive data? Well, I think to our conversation, maybe especially so this week during summit, it's now evolved into, how do you prevent unauthorized access, right? And so you actually block attempts that are not intended for sensitive data before it even happens. And so that goes back to, how do you think about cyber resilience, going back to defense and depth? And it really starts with identity now as a security perimeter and identity as a control plane, which is who the actor is, right? Who authorized that actor? And then, of course, now going to the data side, which is, what is that actor now allowed to do with that sensitive data, whether that actor's a human or an AI agent?

Dave Vellante

>> So what I've heard from a technologist is you got to protect the agent. That's where the focus should be. I've heard, "No, no. You got to protect the data.2 What you're saying is both. It's got to be-

>> Yeah....

Dave Vellante

>> layered approach.

>> And that's actually one of the reasons we're here today is talking about, for example, how our two companies can come together around securing the agent identity, and then, of course, with Snowflake, securing and governing the data access.

Rebecca Knight

>> So, Artin, you were talking a little bit about the different kinds of requirements that customers have, things that they're asking for that they maybe weren't a year ago. How do you describe the ways in which AI workloads have fundamentally changed expectations around performance and ease of use and security?

Artin Avanes

>> I think it starts with accuracy. So really, if you look at our customers who have deployed now agents at scale, it comes always down to how accurate and trustworthy those agents are, right? And there are different dimensions to it. I would say the capabilities we offer through the semantic layer that we introduce are a critical building block to make really AI successful at scale. If you don't have a shared understanding of your data, it's slightly autogonal to security itself, but it is a critical piece when we talk about productivity and accuracy of these agents. The semantic layer is a critical piece to it. And we announced at summit additional capabilities, like Horizon Context, that really taps into metadata that sits outside of Snowflake to also, again, power your agentic workflows and processes much more accurately moving forward. So I think that is really an important pre-req for our customers to actually deploy agents and AI applications at scale with success.

Dave Vellante

>> Do you tap into that, Nancy? Like, what parts of the Snowflake stack matter to 1Password? What are the most salient aspects of it that-

>> Yeah....

Dave Vellante

>> have value?

>> So this goes back to our ethos, which is as the vault for the internet, we contain sensitive credentials, things like API keys, tokens, SSH keys, so on, so forth. And so from a data warehousing perspective, Snowflake has incredible amounts of super sensitive and top-secret data for many enterprises out there, including 1Password. And for many of our customers who also use Snowflake, the question becomes, well, who is that agent? Who delegated authority to that agent to be able to access Snowflake to, let's say, create an analytics dashboard or be able to answer important business questions? And so that goes back to the originating identity of that agent. And so there are ways that, for example, we can bind intent cryptographically to that agent identity such that as that identity is issued to the reasoning layer, to now the data access layer, that intent stays pristine. And you don't have drift from the original intent for which why this agent was created in the first place. And so that's the work that I would say alluding to what we're doing together, and it's so important. Because going back to your earlier question, it's really a case where identity security now meets data security, and you have to preserve that intent across each layer of the stack.

Artin Avanes

>> I would say accuracy was one piece where the semantic layer comes in the picture. Now, with respect to how we integrate with partners like 1Password, it's really around our start of the notion of identity. I mean, everyone is in the industry trying to figure out the story around agent identity, and we introduce at summit the notion of an agent identity that helps our users to differentiate between a human user and a actual agent user. So that's where we can then work very closely with the broader security ecosystem to really secure agents and what they can do. Because the notion of an agent identity is just a building block. With an agent identity, then you can combine it with access policies or security policies that allow you to protect your sensitive data and allows you to define what an agent can do or cannot do relative to a human.

Dave Vellante

>> So agents are onboarded and offboarded a lot faster than humans are. So how do you think about handling that dynamic?

Artin Avanes

>> It comes down to a couple things. First of all, I would say there's still a friction, specifically when you think about highly regulated customers on onboarding of agents. So it really depends on the customer vertical that you're talking about. There's still a lot of friction when it comes to, can I deploy hundreds or even thousands of agents seamlessly? And we are trying with our new AI security capabilities to accelerate that emotion. So really about how can you build that trust to actually allow customers to deploy at greater scale. Now let's talk about the customers that actually have already deployed and might have gone through that motion fairly quickly. They come back and say, "Well, I want now to better understand what's going on. Who is doing what?" So that's why we also introduced our extension to Trust Center, which is our security product within Snowflake, which we announced in Christian's keynote, which is around the security posture for all of your agents that gives you visibility of what agents are running, under what role, under what permissions, what scope they have. Are there potential violations or warnings that we should surface to you automatically? So these are, I would say, the different spectrums from highly regulated ones where the onboarding-offboarding is actually not easy towards the ones who actually have onboarded very quickly, but then come back and say, "Wait, I want to actually manage and govern that better then."

Dave Vellante

>> But really different paradigm than human-

Artin Avanes

>> For sure....

Dave Vellante

>> security versus...

Artin Avanes

>> For sure. And that's why I'm saying the ones that are highly regulated are concerned about, well, now you can easily deploy hundreds or thousands of agents. It's not anymore as slower with the humans, but that's where the friction comes in the picture. And that's what we are trying now to help accelerate with our new AI security capable.

Dave Vellante

>> What are you telling customers about PQC, post-quantum cryptography? Is it something they should be paying attention to?

Artin Avanes

>> I would say yes. The question is always the time horizon. Is it something that would be extremely relevant short term? I would say probably not. Versus long term, definitely yes. So it's really about more around the time horizon than-

Dave Vellante

>> And I think it also-

Artin Avanes

>> Oh, go ahead....

>> relates to where that cryptography is put into place. So for example, as 1Password, we take our mission for protecting credentials very seriously. And so that's where maybe we'll come up with a recent white paper on how we're going to future-proof our vaults or vault encryption against future quantum.

Dave Vellante

>> Because people, hackers, will exfiltrate cryptographic or data that's encrypted with the hopes that someday they'll be able to hack it with quantum, even though it might not be here for five years or however it is. So in your white paper, you talk about how you're essentially solving for that now with the expectation that at some point it's going to be exfiltrated. Is that right?

Dave Vellante

>> Future-proofing, because... you know, and won't name names, but certainly in the industry, right? There have been cases where vaults have been, let's say, brute-force spammed and sensitive credentials get leaked. And so that's one thing that we're super proud of internally, which is we have a zero-knowledge architecture that we design or put all of our products through. And what that means is, even if we were to get breached, an attacker would never get access to sensitive credentials that are stored within our vault.

Dave Vellante

>> Yeah. So you can't even get to them, right?

Dave Vellante

>> We can't even see them. That's actually what we say, right? Attacker can't see them because we can't even see them.

Dave Vellante

>> It's a very AWS mindset from your days there.

Dave Vellante

>> Exactly. We actually built it on AWS Nitro. We're a poster child for-

Dave Vellante

>> Yeah. Okay....

>> Nitro Enclaves.

Dave Vellante

>> Yeah, that makes sense.

Rebecca Knight

>> So, Nancy, what is standing out to you at this summit in terms of announcements that Snowflake has made at a time when AI adoption is increasing so rapidly? What is standing out to you about its approach and its strategy here?

Artin Avanes

>> Yeah. So maybe going back to what Artin said around announcements such as agent observability, different security controls. I think the theme for me that really stood out this week at summit is around the agentic enterprise, right? The conversation has really moved beyond, "Hey, let's secure models." That was really, I think, gen one of AI security. And now it's more agents are here. They're in production, right? We're talking about agent swarms, hundreds, thousands of agents. In fact, I just spoke with agent sandbox company that says they can now spin up concurrently one million-plus different sandboxes. I mean, just think about that scale. And so when you have that scale... And specifically for this customer, they're building this agent swarm to do DevOps work. So if an agent did an update to a database or maybe did a config change, how can you actually map that trust and accountability back to that specific agent in a swarm of millions of agents? And so this goes back to what Artin just said around security controls and observability tying back to accountability.

Dave Vellante

>> Where do you want this relationship to go? What should we expect a year from now we're back at Snowflake Summit 2027. What do you guys want to be able to say that you can't say today?

Artin Avanes

>> I think I would love to be able to cover the entire life cycle from a security perspective. As Nancy was saying, it starts always with the identity and goes then down to the data access. So the entire end-to-end life cycle and building a seamless, very secure integration. As we are bringing the new Snowflake native capabilities, we will always need to have the broader security ecosystem integrate well. Because we can introduce a bunch of native security capabilities. If the ecosystem is not well integrated, there's always risk, right? And I think that is, high level, I would say, the North Star that we want to work with and get also feedback from our partners to say, "Look, this is areas that we need to jointly invest." It's really a joint effort, and it's not one tech partner or vendor solving itself.

Dave Vellante

>> So closing those seams with collaboration and presumably standards that-

Artin Avanes

>> Yes. Yes....

Dave Vellante

>> help you close those gaps.

Artin Avanes

>> And again, the entire life cycle and workflow when it comes to the propagation of the identity down to the data systems like Snowflake.

>> Yeah. I mean, very simple. Maybe one-liner, which is 1Password and Snowflake helps keep your data secure and secures your agent access to data. And this goes back to, I think, Snowflake, you're one of the most innovative customers that we have, which is... I heard from Sridhar, actually, during HumanX conference that you've already built skills using our 1Password developer tools.

Dave Vellante

>> Yeah.

>> And so I think that forward-thinking around the agentic enterprise and everything that you've launched at summit carries through some of the security work that we're about to do together.

Rebecca Knight

>> Excellent. Well, Nancy and Artin, thank you both so much for the conversation. Really interesting.

Dave Vellante

>> Good to have you. Thanks.

Artin Avanes

>> Thank you. Thank you for having us.

Dave Vellante

>> Thanks for having us.

Dave Vellante

>> You bet.

Rebecca Knight

>> I'm Rebecca Knight for Dave Vellante. Stay tuned for more of theCUBE's live coverage of the Snowflake Summit. You're watching theCUBE, the leader in enterprise tech news and analysis.