Name: Govern the Data Powering AI with Cyera and Snowflake
Uploaded: 2026-10-20T16:31:31.469Z
Duration: 44 min 36 s
Description: Govern the Data Powering AI with Cyera and Snowflake

Transcript for "Govern the Data Powering AI with Cyera and Snowflake": Alright. We'll give it another minute or so for folks to hop in. Hope everyone's doing well. Got folks from Winnipeg, Midwest, New York, North Carolina, Saskatchewan. Awesome. We have people from everywhere. Yep. From India as well. Alright. Well, why don't we why don't we start kicking it off? Welcome, everyone. Thanks for joining us today. My name is Michelle Zhang, and I lead, Tech Alliance's product marketing here at Saira. I'll be your host for the day. And today, I'm super excited about this topic. So we're gonna be talking about how to govern the data, powering AI, whether it's securing the data that agents, can access or the agents securing the agents themselves, you know, across Snowflake environments. So I'm really excited to introduce our two speakers. First is Gal Pollock. Gal is the senior product manager here at Sayera, and she leads a lot of our joint integration development with Snowflake, including a recent, integration on access governance directly in Snowflake. And then our special guest speaker is Navneet Shukla. So he is our principal AI, architect, for Cortex at Snowflake. In his role, he leads product integrations across Snowflake and Sayre as well. So to... Together, they're gonna walk you through how Sayre and Snowflake are helping teams, you know, move faster with AI, while keeping their data and their agents trusted and secure. And so we'll be walking you through use cases, you know, customer stories, and then also a demo at the end. So, Galen of Nate, over to you. Thank you, Michelle, you. for the introduction. Alright. Let's kick it off. So I want to start, with painting a picture a little bit, of what's fundamentally changed in security, in the past few years. So for years, the security model was built around the perimeter. You protect the network. You lock down endpoints and secure infrastructure, and that was the playbook. But AI is now opening up new exciting possibilities. It has done so for the past few years. It will continue to do so. It it can really help people find insights and data faster, And it can make information easier to work with and bring new capabilities into everyday workflows. And to do that, AI systems across, can access content, across the entire environment. They can interpret context and operate across different environments very, very fast. So it creates a big opportunity, but it also makes, makes it necessary to rethink how we secure and make most of our data. So if your network is the walls of a building, for example, AI can move straight into the filing cabinets. So security can start closer to what matters most, which is the data itself. And a data first approach helps organizations adopt AI with confidence and unlock more value from their data. And that shift from perimeter first to data first security, is the foundation of everything we're going to talk about today. But before we dive deeper into security, I want, to invite my peer and colleague Navneet from Snowflake to share how he sees these opportunities come to life for Snowflake customers. So Navneet, take it away. Yeah. Thank you, Gail. Hello, everyone. This is Navneet. Good morning, good afternoon, good evening. So as we all know, right, so AI is transferring, transforming how we used to work and actually in how we run our business with the specialized agents such as the client, intelligence agents for the sales or audience and campaign agent for the marketing. And I have worked on the various and forecasting... Building forecasting agents. So those... What we see, like, is how we operate on a day to day basis where we are kind of trying to create specialized agents. And this specialized is it's it's kind of supplementing our work and helping us to be be more proactive. On the other side is we see how business are running as well. So there we are creating the specialized agents such as a growth dip, growth, deposit growth for the banking. We see this claim in prior authorizations for health care. And then in the retail, we see this merchandise decisions, which kind of telling how you can able to go and build your retail application altogether. With all these experience of growth and in in terms of, agents, we see this kind of bring a new challenges as well. And before we go and challenge, what we're able to go and, do is, we see the evaluations, how the agent has has explodes. When we started in 2024, it looks like, it's just a two year, but it's look like this thing passed where we'll be building the chatbots. And when we have started building chatbot, we started encountering all the challenges around the prompt injections, etcetera. Right? And thereafter, once we have twenty twenty five, when we have moved forward, which where we see there is a MCP have arrived and we start connecting all different ecosystems with your data, CRMs, and other applications using the MCP. And what we see here now with the MCP and all those g g n t applications come into pictures, we're starting explosion of growth of the agent. And what we have projected is in 2025... In 2026 or or, like, you know, we see this growth growth of, agent. And there's a projection in 2028. We're looking to have almost around 150 thousands agents being built across multiple organizations or fortune 500 that will be building around 150 thousands agents there as well. So when you increase the number of agents and when you increase the number of connections around the agents using the tools or using the MCP or using any other services, it's creating IT and security risk. What we see in our customer is they've been building, the AI agent, which actually surpassing this security and governance. And that's where we see, like, you know, how you can able to go and monitor those agents which is being built around. And when you're building all these agents in rapid scale, we have seen, like, this creating statistics problem, which we're looking to go and talk in our next slides. The first problem which we see when you're building all these agents is around the runway AI strength. You must have seen those, famous, LinkedIn post from one of the Fortune 500 companies where they said they outpost, outpaced their entire yearly budget of the AI in just four month. Other one is the security incident. What we see here is 88% increase in in the enterprise security around the agents where if they see, like, you know, agent been accessing or deleting the data, or it's providing the wrong information altogether. And the third one is the shadow AI and the compliance liability. As people or as organizations providing these tools to their employee, and there are set of tools which is available across industry for you to go and use it, many times companies comes to us, our organizations comes to us, and they do not even know who and what kind of tool they are using. So they just wanted to know who using the tool, what kind of tool is being used, when the tool is being used, and how much is it costing around the tools. Right? So this way, they're not able to go and capture, around what all things is being used in their organization, but also they're able to go and manage. So able... They're able to go put the budgets and everything as such. And that's what we try to do on the Snowflake as well. We provide two set of services and tools for you to not only build a better agents, but also able to go and monitor it. So if you see here, this has been kind of divided into three different sections. First one is you can able to go and use the hardnesses such as Snowflake's co work or Coco, which allow you to go and perform your, day to day work. With the co work, the business or the business people able to go use the natural language to ask the questions on the data settings across the Snowflake and external system using the MCP connector with a co work developer able to go and use it and build their, data and, AI pipeline altogether. After that, you will go and access the Cortex agents. The Cortex agents able to go and stitch together all the data, which is there into your snowflake, whether it's a structured data, whether it's unstructured data. And then you're able to go and use the codecs that I mentioned is the... The job of the codecs that I mentioned is to perform the difficult task in just a single SQL statement. For instance, you can able to go and use summarize, etcetera. And then you will be have... Providing the entire agenting ML experience altogether for you to go and build a... In the agent as well, which comes with, you know, you can have a a OTAX SDKs, which allow you to go and build the end to end agenting solution within the Snowflake. And thereafter, when we talk about the codecs, the context clear, this is where we are providing, tools such as the codec sense runtime, which allow you to go and connect all the data and not only the data which is sitting within the Snowflake, but also if you have a data sitting outside of Snowflake and if you're looking to go bring the data. Remember I told, that the slide where the CRMs and all was there. So if you have a data which kind of bring from different data sources, why are you using MCP or why are you using APIs, etcetera, you can able to go and use the codecs as kind of give you an idea where all the data resides. The semantic view where you can able to go and define the data, if the data setting is in the snowflake, the structured data, you can able to go define the data, how the data look like. You can able to go create a description around your data. You can tell how the data look like, and, also, you can able to go and put a business logic. What will happen with this one... But this will allow you to have your agent to have a far better context and provide you an options. This is very broad label. We offer far more capabilities. Other one is where I primarily work around the Kotex AI gateway and Kotex inference, where not only you're able to go and access the model, provided by the Frontier, labs such as OpenAI or Anthropic, but you're also able to go and access the OSS model as well, such as, GLM, QEMI, GROG, other model which you can able to go and and access it. Other ways, we just, not only you're able to go and access those model, via first party model provider, you can able to go and access those model through, through the third party model provider such as a Bedrock, a Google, Azure, etcetera. Why it is important? So this way you can able to go have a business continue... Continuity and a and a compliance requirement met. So this is a new, new, service which we have brought in. We call it... Which is in public preview. We call it Codex AI Gateway. The Codex AI Gateway allow you to have a centralized place for you to go and manage on... Manage and govern all your agents and MCP and tools. So it it can have particular four pillars where we are having the model optimizations, which you can allow you to have a intelligent model routing, which take account of latency and accuracy. The second one is the cost governance. In the cost governance, possibly, we will be... Allow you to go and track your cost. You will be tracking the cost per span of your, of your agent, and you can able to go group group together to per sessions as well, and then you can able to go and group together per 10 as well. So if you're building the multi tenant agents, you can able to go and see how the... How that agent will be costing across multiple spends and turns altogether. Thereafter, you've been able to have a security where we allow you to have a RBAC, which you can allow you to go and access, create an access and governance around the model, who are able to go and access the model, and how they're able to go and access the model, and what kind of a data they have access to on this model. Also, we are providing the trace table, where we allow you to do the real time monitoring, and then you can able to go and see the performance metrics as well. With that, I will hand over to Gail and talk about how we, as a Snowflake and, CRO working together to, supplement the capabilities offered by both the platforms. Thank you, Navneet. And that is truly amazing, all the capabilities that we see, and, it really scales very quickly. So it's truly amazing, what can be done today. So, I'll talk now how Sarah fits here, in helping organizations and enterprises adopt AI securely, especially in Snowflake. So, I'll start with the Sayer platform. So, this is the Sayer platform. It's what we call the trust layer for AgenTek Enterprise, and there are three security pillars here that we have. So the first one is our data security. So that, is our DSPM and our DLP, and it's the ability to understand what sensitive data you have and governing how this data moves. The second pillar is our identity security. It's understanding every identity and what access this identity has, and whether that access is to sensitive data, or if it's a wide access, third party access, etcetera. And the third one is our agent security. So securing a... AI agents and the actions that they take. And across all three pillars, we enforce continuous trust. That means less privileged access, runtime enforcement, comparing behavior against intent, and full audit trails and compliance. And these aren't separate tools. It's one unified platform that gives you visibility and control across data, identities, and agents. So what does trust layer actually means? Here's the way to think about it. As enterprises deploy more AI agents, they need a foundation that understands what data exists, who can access it, and what agents are doing with it. And that foundation has to work across your entire environment, not just one cloud or one data store. And, it has to give you the following capabilities. So you need to be able to discover, what data do you have. You need the ability to govern who can access or who should access this data. You also need the ability to protect it, and understand how you secure it. And, you have to be able to validate, if this this data is being used correctly, this agent is behaving correctly. And at the end, you have the result of having the ability to enable, the organization to adopt AI. How how can you make this fast and safe at the same time? So that's what Sierra does, and Sierra is the layer that makes AI trustworthy at scale. Now if we're looking at how Snowflake, comes, into the picture, Salesforce Snowflake helps customers secure the data foundation behind AI. So it gives teams visibility into the sensitive data in Snowflake, access pass, identities, roles, misconfigurations, and risks across Snowflake environments. And there are three parts here. So, the first one is the speed and scale. SARE is built to discover data at Snowflake scale. And when I'm talking about Snowflake scale, it can be petabyte large environments being scanned overnight. We already saw exabyte large environments in Snowflake customers. So it's really a tool that is built for scale, and Sierra has adapted, that ability to, to scale and speed, which really matters here. And it mainly matters because AI projects do not wait for slow security programs. If the business is moving quickly, security has to keep up. The second thing is, the precision. So AI, the AI native engine Sayra has, learns the data types that are unique to each business. And I'm just gonna give a number here, that 86% of our classifications are learned data types, which means that, Sierra has matched this data classes to be unique to the business that, they belong in. And Sierra classifies the sensitive data with 95% accuracy. So that means that customers can find the data that actually matters and not just the data that matches matches a generic pattern or a regex. And last, the action. So Sayuri helps customers move from discovery to remediation inside Snowflake, and that includes understanding sensitive data at the column level and applying Snowflake tags, that support dynamic masking policies, identifying excessive access, and helping teams reduce real risk. And this is what makes Sayera for Snowflake so powerful. Snowflake helps organizations activate data for AI, and we've seen this happen, in wonderful ways. And then Sayero helps make sure that this data is understood, governed, and protected before AI risk scales. So here at Sierra, we call this actionable data intelligence. So, Navneet, how how do you see the importance of having actionable data intelligence for Snowflake users? it does like, one is when when we talk about our partnership, right, with with the CRRs. And, that's where we see the complimentary capabilities offered by by both the platform. And, and when we when we talk about making the data intelligent, I think we surpassed that, like, okay, you should have the lineage and the data quality and everything as such. Now we are looking to have far more, granular insights around the data. And, and I think, that's where the, Sierra brings that unique capability is where it able to go provide the data visibility. It able to go provide the data classifications. It able to go and provide the risk prioritizations. It allow you to go have the policy, enforcement, on this on this, Snowflake platform, which kind of, you know, provide the governance and data foundations. With all these capabilities, now you can able to go and trust your data. So when you will be building and creating an AI agent, and then you would be looking to go and see the h I... AI agent is not able to up to par or it's not providing the right output or hallucinating in a way, you can able to go and trace it why this has been happening. So that's what I see, like, a differentiator of the Sierra platform on the Snowflake, and I'm super excited for our partnership and what we're bringing together. Yeah. That's beautifully said and, echoing, what, Navaneet said here. The goal is not just to discover data, or classify the data. It's to really help Snowflake, users and organizations understand data well enough to make decisions and take actions. And Sarah addresses this with our actionable data intelligence, mindset. So if I'm gonna speak on different, use cases that, Sarah offers for Snowflake. Let's see how it looks, in practice. For our BSPM for Snowflake, Sarah scans, your whole Snowflake environment and builds a complete map of sensitive data down to the column level. It identifies which tables contain regulated data data like PII or PHI or financial records or, specific, IPs, that are unique to your organization. And then it discovers business critical data that's, that's important, to understand and have visibility for. And from there, it understands the access paths, which user roles and AI agents can reach this data. All of this happens automatically at scale without disrupting your SnellSafe workloads, And the result is a living inventory of your data landscape and that updates continuously as data changes, when new tables are created and, access patterns evolve. Now once they understand your data, you can act up on it. This is what we call Snowflake access governance. You can apply Snowflake tags automatically based on what's their classifies. And you can create, dynamic masking policies. And dynamic masking policies is a very powerful Snowflake native feature. And once this dynamic masking policy is attached to a tag, once, Sayera tags the sensitive columns that Sayera found, then automatically, this, dynamic masking policies are applied to that column. So you can find overprivileged roles and write access to this privilege, and you can monitor AI agents, see exactly what data they're touching and whether that matches their intended purpose. And this is where intelligence becomes action. So it's not just alert. Actual remediation workflows reduce risk in Snowflake. Now as the landscape evolved, Sayre and Snowflake partnership, through the years evolved as well. And, we've recently announced, AgentGuardians, a few months ago, and this adds another layer of support for the emerging need of scaling AI safely. So let me introduce agent guardian for Snowflake. Agent guardian gives you four critical capabilities. So first, you have inventory... An inventory of every Cortex AI agent and workflow. So that gives you the ability to see exactly which agents exist, what they're doing in Flag Shadow AI. The second thing is you can map each agent's full data reach. You can understand the complete access path from agent to data in Snowflake. Third thing is that you can serve this implicit permission risk. Agent Guardian identifies where agents have access that they shouldn't have, and it lets you enforce, risk based policies. And last, you have one click remediations. So risky agent access to data and Snowflake gets fixed with with a single click. So there's no need for a manual workflow, or specific engineering tickets here. You can just do that, from Sierra through Snowflake. And this is how you can scale your AI agents on Snowflake without scaling risk. Agent Guardian really gives you visibility, control, and fast remediation for every Cortex AI agent in your environment. So this is a great up... Upgrade time to, stop for a short demo, and I'll share my screen here. So I want to begin with, showing, the DSPM capabilities, with our, Snowflake access governance. And if we're going to issues and looking at Snowflake issues, we can see, for example, an issue here for credit card numbers and plain text. Now if we go into, this issue... So first of all, we have, a summary of what is this issue and what makes it critical. Why is this a risk that I should, remediate. I can also see the affected data, that, has and includes this risk, and we can really see that there are credit cards here living in plain text. And from the Sayer platform, we can go ahead and start a remediation of applying a Snowflake tag. Now if this tag is attached to a dynamic masking policy, in Snowflake, once we choose that tag and apply it, all the columns containing, credit card numbers will automatically get dynamically masked, and this really reduces the risk at the click of a button. And if we're looking at our agent guardian, page for AI asset, we can see that there are a couple of Snowflake Cortex agents here. And for these agents, we can have a full map of the agent intent. We can see the risk, that we found for this agent, the agent graph, and what it can access activity, for the last period of time. And this is really interesting because we can also see all the, data classes that this agent can access. So if we're seeing an agent, that is used for, financial services, it maybe shouldn't have access to, specific data, classifications that a financial services agent, should have. And this is, an amazing way of displaying the full cycle of the Sierra and Snowflake solution. So continuing, to a real example of what this looks like in practice, and I'm bringing here a customer story, which was, recently published as well. So we had a UK fintech, company that brought its data state across AWS and Snowflake, into one ClearView in Sierra. And having this ClearView helped the team... The security team of this organization reduce sensitive data instances from a 170,000,000 records at risk found to, just five, 500 thousands. And this is a risk reduction of an amazing skill. It... Sayer brought visibility across Snowflake and the rest of the environment, and they could prevent personal data from flowing out of their core system into other databases and data sources. They also got ahead of data minimization expectations, and the industry standards. So for a financial services company, that matters because trust is everything. And as the customer put it in their own words, partnering with SEER was key to building and continuing to maintain that trust. And I really love that customer story. Right. We have reached the questions, for, our webinar. Yeah. Great job, Gal and Navneet. So, yeah, we've got a couple of questions and, you know, continue adding them to the q and a or the chat, if you have any that pop up. Alright. Let's take the first one, just in the chat. And I think this speaks to the case study that you just shared, as well. So can you speak to any regulatory compliance capabilities or considerations for those of us in highly regulated industries? Of course. So Sierra has, also, compliance frame... Frameworks in the in the platform. We really have a guide through of each compliance, individually. So if it's, GDPR, HIPAA, or other heavily regulated, fields and businesses, We have a set of policies, and specific guidelines, to ensure that the data is compliant to, that that compliance framework. So it's very structured, very easy to ingest in Sayra. And for additional internal policies that the company might have, we also have tailor made policies that can easily be created, by the security teams themselves to ensure that their own company policies are also, reflected in Sayera's, Sayera's findings. Yeah. I think what's, you know, really unique there is, the ability to autonomously, you know, check to see that, your organization is remaining compliant, you know, whether it's the policies that are updating or, you know, within your enterprise. There's a new situation as well. So, it's that autonomous ability to, ensure that you continue to maintain compliance. And then we have another question around reporting. So can you comment on the SIR reporting capabilities? Part two to that question is, can I configure the reporting dashboard to meet our specific requirements or preferences? Yes. So we have both customer reportings, reporting being able to be built in the platform. And we also have our amazing AI assistant, Sai, which can build reports, from natural, language prompts. So, if you have a specific report that, your leadership or your peers would want to see, you can go to SAI. Let them know what reports and what insights you're looking for, and so I would build that report for you. Awesome. Alright, Navneet. We've got a question for you, and this one came through a q and a, so I'm just gonna share it onto the screen. So how can AI agents transform traditional business workflows and improve productivity across departments such as sales, marketing, and finance? Oh, you might be on mute. My apologies. I was speaking on mute. So thank you, Abin, for the question here. This is an awesome question, actually. And we do the walk to talk, and in the Snowflake itself, we have created something called... I'm not sure whether I supposed to name it or not, but, anyways, so, what that it allow us to do is it allow us to go and join all the data together, whether we're data sitting in the CRM system, whether we are getting the data from our marketing team or marketing system, and then we are having the sales data which is sitting across different places, altogether. And. what it allow us to do is, using the codecs agent. Remember I showed that the slide where we talked about, the entire ecosystems, for you to go and access the data from the harnesses such as a cohort where our, marketing team and sales team, they've been asking question using the natural language. And under the hood, what we have did is we have built the multiple agents, which connecting those data through the MCPs and everything as such, and then creating it. Other important part which, wanted to highlight is I think this is kind of a now, defector, like, everyone is building in a similar way. The one thing which is important and when... Where I see the differentiator is the trust factor, which we kind of... Has highlighted during his, his slide and the customer stories as well. Unless until you have not have a 100% trust or you're not able to go and trust it, and especially when you're working with the finance use cases. I was working with the one of customers, where we were helping to build the variance analysis, tool. And when you work on the variance analysis, you need to ensure whatever you will be reporting, which goes directly to the office of CFO. It's 100% deterministic in an output. And that's where it's very, very important when you build... Building any, workflow of your business. Are you able to go and trace it, at the core? Okay? The data which is being provided by the agent, You're able to go trace it. You're able to ensure the data which you are getting are trustworthy. And more importantly, you're able to go and put all kind of a garden around it. Hope I answered that questions. Yeah. And I think there's a a great part part two question here. So I'll share this on the screen. You know, Galar and Navneet, feel free to address this. This goes back to, you know, the prior question of, you know, now different teams like sales, marketing, HR, various levels of regulations required. They all, you know, those teams are now adopting agents and accessing data that they never would have been able to access before, because those agents have, you know, the capability to to find more than a human can. And, actually, in Sayre, we have this, uncovered the statistic that, humans can only access 4% of what agents are accessing. And so what are the issues with surrounding the adoption of AI early? You know? These teams wanna make sure that, you know, they're able to adopt early, you know, and and gain more, efficiency, but, you know, that also can come with, having issues with incorrect information or results or, folks that are accessing information that they shouldn't. So how do we instill trust for AI going forward? Maybe I can take a jab here, so... And then I. will pass to that. So at least one thing which I see here is excess of information. There's too much excess of information, nowadays. And, what I have seen the first hand is not creating the information. It's so much easy. Right? You... Anyone can able to go and write a prompt and then create those informations. And, when I, went on the Snowflake site itself for the... Even even for the internal or an external agents, which we are, which we are, building. The challenge which we see here is, okay, you have... For the for the same information, for the same piece of information which you're looking, you have a multiple documentations sitting across multiple, you know, business units or sitting across with the multiple people. The way you... I think of, like, you know, when you're looking to go and work on how you're able to go and create a... Install a trust on the AI and AI agent altogether is providing the right and meaningful information. And that's where I see, like, because of all the AI slopes, which we are encountering on our day to day basis, it's create a plethora of of information. And sometimes, for you to go and make a meaningful, AI agent altogether, you have to just provide very limited meaningful information instead of just feeding all the data. And I... As an architect, when I see anyone building an agent and just providing all the plethora of information to the agent, they are setting, like, that agent to fail. And more importantly, it's going to be an expensive, It's going to be create the, challenges around the cost and governance and then data classification altogether because now you have so much of data. Classifying those data in itself is is going to be a big task. Now I'll, love to hear guys' feedback, guys' thought on this one. Yeah. That's very well put. And I I also want to emphasize that adopting AI early, for me, I I see that as a huge opportunity. And, I highly encourage, organizations to adopt AI, do it early, try and see which tools, enhance, the work and productivity of the organization. I will say that, to adopt AI, securely and not, kind of, like, create, an enormous rise in risk, I would really recommend really understanding the data, underlying, in the environment. So, Michelle mentioned this, but, AI tools, AI agents, have very wide access that sometimes is almost invisible. The, way that agents also use this access is is pretty exhaustively. So AI agents can use a 100% of their, access permissions, whether humans barely use most of it. So, we really have this exploding risk, that comes with adopting, AI tools and AI agents. But once the organization can really understands... Understand the data, understand where the sensitive data lies, and whether or not an AI agent has access to it, And then from there, make the informed decision of, whether the agents accessing the sensitive data should access it or shouldn't and enforcing these policies. It really kind of, like, bridges that, potential risk, from adopting AI early, and it just enables the organization to work, with AI, which really enhances enhances the day to day life, the productivity, the capabilities that it can bring. Yeah. That was. a good question. Alright. Well... Yeah. Absolutely. Why don't we just take one more question? And, Gal, I think this one will be for you. So I'm gonna share that on the screen. So Scott Good asked, how do we automatically apply Snowflake dynamic data masking in row level security policies based on the real time data classifications that could be found by Sierra? So that's also an amazing question. So Snowflake has this amazing capabilities of, dynamic masking policies, and I've seen this in play, multiple times. It's an amazing feature that, takes access governance to a whole new level. And based on the role that access the data, the data will be visible, in, different, layers of, of encryption, masking, hashing. And so when you take this capability that Snowflake offers and you match them into Sayer's data classifications, you have, you have this, masking policies that are based on, data sensitivity or data classifications. So, the way that you can, practically, do it, you can go to Snowflake, see the data classifications. That's, you can go to a Sarah. Apologies. See the data classifications Sarah, has found. And then from there, with the, click of apply Snowflake tags like I showed in the demo, you can go ahead and apply a Snowflake tag that, has a dynamic masking policy attached. And then, that would automate the whole process of, understanding what data classifications need to be, dynamically masked, or encrypted, and then from there, enforcing that. Awesome. Alright. I think that's gonna be it for today. Thank you, again, Gala Navneet, for, taking the time to walk us through that. And thank you all for joining. If you would like to learn more about our Snowflake entire partnership and of our use... Joint use cases, there's a QR code there. You can also book a demo. And then, be sure to check out the docs section as well. We've got several assets, for you to refer to as well. And then lastly, we've got our upcoming user conference, in November. So 11/09/2010. So we'll be sure to, it's called DataSec AI. We'll be having, you know, a a a snowflake insider session as well. So a lot of great speakers coming up. So we'll be sure to get that registration link out to you. If you're interested, definitely join us there. Alright. Thank you all so much. Okay. Thank you. And take care.