The Context Layer for Your Microsoft Fabric Estate

Razi Shafiq

0:00 All right. Thank you everyone for joining. We'll give it a minute or two to let everyone trickle in, and then we'll get started.

0:15 All right.

1:20 Thank you everyone for joining another Atlan in Action. And we have a great one in store today. We're joined by the Microsoft Fabric team here. And so today what we're gonna talk about is the Atlan and Fabric partnership, where we are today, and how we are continuing to evolve that partnership. I'm joined by Chandru our director of product here at Atlan, as well as Yaron from Microsoft Fabric, who also is a product in product here at Microsoft. What we're gonna do today is go through the partnership how we're building together as well as some demos both on Fabric and in Atlan of that partnership.

2:07 And with that, I'm gonna hand it off to Chandru here to get us started.

Chandru Sugunan

2:14 Hey. Hello, everyone. Good morning, afternoon wherever you're joining from. Welcome to the Atlan in Action webinar series. So let's go. Let's ground on a simple truth. AI is trustworthy as the context behind it. That's what Atlan is building towards. We are building a governed context for your enterprise's data and AI stack. As we proceed just drop your role in the chat. Like, are you on the data leader, architect? You know what is your role? Tell us who you are. So here we go. So governing the data and AI context is easy at small scale.

3:00 Here, you know there is this one team one analytics agent answering these questions like what's your total revenue, you know, where the data lives, you know, who owns these metrics who can write the definition into the prompt, and it just works. Nobody stays there, right? If you look at the right side of the slide, the customer-facing agents, operational agents, and many more such agents in the analytics side, coding tools when you go into the multi-agent platform, that's where it gets things get complicated. And that's where you all are headed into this year.

3:42 So if you, if you see what's actually we are showing here is your system of record are on the left side, your data platforms. For most of you, you know, it is Fabric, OneLake Power BI, Teams, SharePoints shared drives and so on, right? On the other side there is a growing list of agents, and that's where the messy middle in between are. So every team that builds an agent bakes its context into or into that one agent the definitions in the prompts, rules in the code. Then someone builds the second agent, and they redo the same thing. And then the third agent, they do the same thing, encode everything in the same agent.

4:29 So now what happens is the agent one thinks revenue means bookings, and the agent two thinks revenue means ARR. The agent three don't even know what revenue is. And in this case, none of them are broken. The agents are not broken. Each got different context from different teams. So the question here is: how do you centralize and reuse the context you build for agent one and reuse and expand it to agent two and agent three? And every company has, you know, like, its own term for revenue. In your company, what are some, you know, one or two terms that have more than, you know, two or more different meanings to it?

5:10 I'm pretty sure in our case you know revenue means slightly different for marketing versus in comparison to sales. So let's ground what a governed context is. Like, if you ask what is context to your folks, you will get different answers. A BI leader hears that context as semantic models. An architect hears that as ontology. Someone else hears as knowledge graph. So if you level up, what does an agent actually need to do, need to know in order to do its job? It needs three things. First, it needs knowledge, a map of the business what exactly revenue means which semantic models it is, is it coming from, the lineage behind it. And then the expertise.

6:01 How a work gets done. A good analyst, a human analyst, checks the seasonality or what's a new product promotion is launched before explaining a dip or a spike in the, you know, like sales. An agent doesn't, unless you tell, write it down, "Okay, this is what a dip or a promotion or a reality or a seasonality means." And then third is the norms. What's allowed? Who's asking? What they can see. A CFO versus a store manager asking the same question, what answer they get. And then governed here means each one is written once, owned, versioned, and every human and agent see the same thing.

6:41 So that's what, that's what we are talking about in this slide. So how do we centralize this governed context? You know, the whole messy middle showed earlier, right? Context is mined from across the enterprise, governed, served to every agent. Look at the three context types in the middle. The AI-ready data, this is where your Fabric lakehouse, the Power BI models, the reports, transformations, one map of lineage tied to the business concepts like your glossaries, that's the foundation. And then comes the semantics. As you build analyst... Analytics agents, this becomes a big one. If you ask, "Okay, what is revenue?" In ten different ways, you get the same answer.

7:26 On Fabric, much of this already is in your Power BI semantics and DAX measures. And then finally, the skills. This is the newest piece in the puzzle. How works actually gets done in form of an agent can use. And it all goes out to wherever the agent is running, could be Microsoft Copilot or Copilot in Power BI or Foundry or wherever, it does— it doesn't really matter, right? The Atlan's NCP server already works with Copilot Studio, and this is where the, you know, the context is defined once, and wherever the agent is, the context is served in a governed fashion. So if you see nobody builds this all at once.

8:09 We see customers move through, you know, these four different stages. The line at the bottom is the most important thing and, you know, like, you should not skip ahead to have like a scaled motion for your governed context for the agents. The stage one, which is the AI-ready data a human or agent find data they can trust, understand, and use. The stage two is the business meaning, the semantics, the metrics on top of it, the whole business ontology is pulled out out of the, out of the data. You can, you know, you can use it and again, in Fabric, this is the semantic models and the DAX measures.

8:49 The stage three is the agentic context, you know, the tool calling, what is the right procedure and authority, the process behind it, right? The SOPs. And finally, the self-improving context. What agents do in production feeds back without, you know, quietly, you know, failing it, the evals, the feedback loop. Most of our Fabric customers, our mutual Fabric customers, are solid in, you know in one of these stages here, and that's exactly we are going to see a demo from Razi later. And finally you know, for the Fabcon, we released a bunch of things.

Chandru Sugunan

9:31 The integration between Fabric and Atlan is for an end-to-end data and AI governance natively integrated into Microsoft Fabric. Fabric and OneLake are the unified foundation with governance in the OneLake catalog. Yaron can take you inside in just a minute. We extend that context across everything around it, around the Fabric, around the Microsoft ecosystem. The source system before Fabric, like SAP or Salesforce or wherever it is. The other data and AI systems next to it. We get everything into the one broader umbrella. If you remember, I mentioned revenue in one lineage graph. It shows right from the sources, you know, around the Microsoft ecosystem, and it carries all the way across Microsoft eco-ecosystem and beyond.

10:23 So without further ado, I am going to... Oh, I have one last slide. So see, this is the words from the Microsoft leaders at Fabcon. You can see Arun Ulag, who is the EVP of Microsoft, at stage announcing Fabric for... A-announcing Atlan for as a governance layer for the entire data estate, including Fabric, and in the right, you can quote, you can see the quote from Deepti. So with that Yaron, take it away for you know, your section of what's inside OneLake.

Yaron Canari

11:10 All right. Yeah, thank you, Chandru and everyone. Let me share my screen, and we will start. Yeah, I hope you can see the presentation deck. So hello, everyone. My name is Yaron Canari. I'm a product manager lead in the OneLake catalog within Microsoft, and we will give you a very quick overview of OneLake catalog and also how we are open and how we integrate with Atlan and, and other tools and engines that can connect to the data and metadata. So if you're a Fabric customer, or you have heard about Fabric, you better know this slide. We have in Fabric, we have multiple workloads working all together based on the shared Fabric platform.

11:53 We have the unified Copilot M365 Copilot and the OneLake and Governance at the center. And OneLake is the central data gravity where you can bring data not only from Microsoft data systems like SQL and, and, and other Azure, but also from other many connectors outside, including Snowflake, Databricks AWS and, and much more. So you bring data to OneLake, whether it's like with the data pipeline, data flow, but also with mirroring and shortcuts, you can just bring the data with zero ETL processes.

12:34 So everything is within OneLake then, and then you can do data engineering, data scientist, data analytics based on your OneLake information. And OneLake or the OneLake catalog is also integrated to the Microsoft ecosystem through many applications, whether it's for business users within Excel or AI developers and agent through Copilot Studio Microsoft Foundry Teams and, and and many more. So bringing the data in, then integrate it to the Microsoft ecosystem and to third-party integrations as well. We can see that soon.

13:17 And OneLake catalog, whoever didn't hear about it yet, this is a service within Fabric within the Fabric platform. This is in GA already for more than a year now. And it's, it's broadly used. So we are integrated with the OneLake catalog across many experiences, whether it's within Fabric or outside of Fabrics, within Teams, Excel, Foundry, and more. It's broadly used by more than two hundred and forty K customers, organizations already. Whoever is using any workload in Fabric, whether it's Power BI, data engineering, any other workload gets out of the box.

13:59 The OneLake catalog is a service, and we have more than three hundred and fifty million assets indexed within the OneLake catalog that anyone can use. When we look on the OneLake catalog, we see like we have kind of these three pillars. We have the unified data discovery and curation to let the data curators to add metadata to make data discoverable. We have data governance for admins, whether it's a tenant admin, domain admin, setting up tags and policies. And we have the unified data security to put the OneLake security and the workspace permissions to control who can access the metadata and data.

14:46 And everything is on top of an open platform. We will touch on that also in the next slides. But we do work with partners, including, of course, Atlan, a strategic partner, and, and others to be able to to read and write metadata from OneLake. Let's try to do a live demo, a very quick one. So I'm sharing my browser now. So this is the OneLake catalog. You can see it today already. And it spans across all the data estate in your tenant across all workspaces within that, across all item types, whether it's native Fabric or mirrored items like Snowflake, Databricks AWS all together within the OneLake catalog.

15:35 Whatever you bring to OneLake, you will see in here. You can filter also by domains if those were defined, like finance domain, for example, or to filter by tags. Then as a... Let's take a perspective of, like, a data curator or data engineer. You can browse through the catalog to the schemas and within the schema to specific schema to a table. Then for the table, you can see some more metadata, including column names, types, comments, and, and other metadata, technical details. And also because this is within Fabric, you can see data preview of the data itself for a table. Let's see. That's coming. Might take a second. Okay. Anyway we will not wait for that.

16:21 But as a data curator, I can also go to the item to the item, to the lakehouse. I can see the lineage. The lineage is for Fabric. Within Fabric you can see either a list or a, a lineage graph, and also to take actions as a data curator. Sorry. Yeah. So you can also take actions and to promote the data. Let's see. Yeah. You can take actions to promote the data or certify it if it meets my organizational standards, to put some tags and, and to do more.

17:05 So this is from the data curator's perspective, but then also like the data analyst or the business users can go to the, the search experience and just to search for a specific whether it's a table or column, and then also to find it through the search experience. You don't need to know where the data exists. You can just find it. So you can see those items, and then you can navigate to them and take additional actions. So this is very briefly for the OneLake catalog overview. And this doesn't stop at Fabric. We integrated with many applications within Microsoft. And the beautiful thing is that we make sure to enforce of course, permissions and security.

17:51 So if a user connects from Excel through the OneLake catalog to the data, it still means that the OneLake security that is applied within the OneLake catalog is still being enforced, so users can connect to only data that they have data access to.

Yaron Canari

18:09 And now with something we introduced just recently and also we will see in the next demos from Atlan that we provide a granular runtime lineage within Fabric. We are starting from Spark. So each and every Spark job execution, we collect the lineage events in open lineage standard format. We store those lineage events within the customer OneLake, and that makes it accessible to the customer or to the, the, to Atlan on behalf of the customer. So anyone can integrate with this data and can build the graph based on this information. So we started with Spark. You will see a quick demo for that.

18:55 But we're also working with other workloads, including data integration pipeline and copy job, and later data flow, to bring the same data in the same format to the same place. So for consumers, for customers, for partners, it will be easily, easily consumed. Okay I hope that you can hear the audio. Let me play that. If you cannot hear the audio, let me know.

20:52 Okay. Yeah, so just I think we have two more slides just to tell. So we, we insist on being open within OneLake, within Fabric. So we have these three layers model. You can think about the storage layer with Delta Lake and Parquet. Then on top, we have the OneLake table API that knows to work with Delta and Iceberg. And the layer number three at the top is the catalog search to be able to find items and sub-item within that. So with these three layers, you can integrate with engines outside, with partners, with customers that can read the data. And we also have the scanner APIs, admin APIs.

21:34 We have multiple integrations, and of course, we will focus on Atlan in in this session, this strategic partnership. And we have the scanner APIs with both service principle and OAuth authentication. And with that, we basically provide metadata and lineage to external tools to integrate with and to build their solution on top. And also for that, we have many customers already using that with either Atlan or oth-others, or they'll build their own solution. So that was it. Thank you very much. I think, Chandru, you are taking that from here.

Razi Shafiq

22:21 Yeah, Razi is taking away.

Yaron Canari

22:23 Thank you.

Razi Shafiq

22:25 Yeah. Tha-thank you for that. And now we're gonna finish it off with a demo here of the additional integrations between Atlan and Microsoft. And with that, I'm gonna start here in Teams in actually Copilot here in Teams, and ask a question. And the question that I'm asking is, I want to list the users in our top five accounts by their email. What this is actually doing is going to both Atlan and Fabric to get this answer. And as you can see here, right away it says, "I can't return the raw email addresses," because that column is tagged as PII in Atlan. But then it does query Fabric directly to get the rest of the answer.

23:16 And so in that way, we're keeping that privacy that has been coded both in Fabric and Atlan, but allowing users to query and ask questions directly from Copilot, whether that be in Teams, through the web, et cetera. With that, let's actually go and look at this specific column. So let me jump into Atlan. So here we can see the email column on our dim user table, and the reason that email was not shown, because we have it tagged as PII with a sensitivity label, letting us know that because this is PII, not everyone should have access to it. How did we actually get here? How did we get to this point?

24:01 So let's, let's talk about that. That's gonna start here in our marketplace and this is where we're gonna have connectors into all these different pieces. So Fabric as an example. We can connect into Fabric and OneLake to start to understand what exists within Fabric, what exists within Power BI, et cetera. Once all of these are connected, we create in Atlan what we call our enterprise data graph, and let me show you one node on that graph. It's gonna be here. And so here we can see a report in Power BI. I'm gonna click into it, our Customer 360 report.

24:44 Here I'm gonna have all of that relevant information about what this report is, how to use it, the audience, et cetera along with lineage, and that lineage is gonna be built through many different pieces. And if I jump into lineage here, I'll just zoom in a little bit. Here we can see the actual Power BI report. I'm gonna open this up all the way upstream. And as I follow this, we can see that there's a semantic model in Fabric, a warehouse table, lakehouse tables. But ultimately, this information started in SAP and Salesforce. From those, it was loaded into Snowflake and through data flow gen two into lakehouse tables here.

25:31 And as well as Yaron showed, we can also show that linked spark job that is moving and transforming that data in between Fabric. Not only that, I am now also able to see column-level lineage, not only within Fabric and Power BI, I'm able to see this column-level lineage all the way back to the source. So here we can see both in SAP, Salesforce, as well as Snowflake, the column-level lineage that exists.

Razi Shafiq

26:06 There was one question that came up in the chat about how can we curate and add descriptions, et cetera, to these objects. Here in context agents in Atlan, I can do just that. So I can bucket my information and have agents run to add curation and documentation. What used to be a, a team of stewards writing this manually, we now have agents that run. So here's an example of one of our agents. Our description agent not only does it look at the schema and the naming conventions, it looks at those transformations through lineage. It will look at query history to start to understand how are teams actually using this data?

26:50 What types of questions are they asking? It will use all of that information to provide rich descriptions, rich readmes, as well as SQL intelligence into what are common joins, filters, and relationships. Ultimately, all this information that's curated in Atlan can be made available natively inside of OneLake, and that's gonna be through our shortcut integration. So here you can see Atlan's lakehouse that has all of that metadata in iceberg tables. We can have a one-click to, to Fabric and to OneLake to create shortcuts, so now all of that metadata is available directly in OneLake.

27:39 And with that, that concludes the Atlan portion of the demo. Please do continue to ask your questions in the chat and we have seen some come in. We'll, we will answer some of those questions now.

Razi Shafiq

27:57 I'm just gonna jump into one of these questions. So one question from Gaurav. "How does OneLake and Atlan complement each other given both offer data discovery and cataloging features?" Chandru, do you wanna take that question?

Chandru Sugunan

28:24 Sorry, say the question again, Razi. I was typing to a different question.

Razi Shafiq

28:28 Yeah. How does OneLake and Atlan complement each other given both offer data discovery and cataloging features?

Chandru Sugunan

28:35 Yeah. That's a great question. I'll start, and Yaron, please feel free to add anything I missed. So as Yaron said, OneLake catalog, if you can imagine everything inside the Fabric world the OneLake catalog will get deeper context around technical metadata, lineage, and whatnot. And if you think about an enterprise stack it goes inside and beyond Fabric. Now you will have, you know, something running in SAP or Salesforce or, you know, even in SQL on-prem.

29:19 And this is where Atlan understands your context, and beyond that, there is business context around glossary terms definitions you know, your, your how do you run the business, the SOPs, the unstructured world, the context around Fabric that make sure that your agents and human are successful with that context. That's where Atlan fits in. Yaron, feel free to... And by the way, we do deep integration with OneLake and Fabric overall. So everything the OneLake catalog is building is available for you know, stitching together in Atlan.

29:58 And once we do that entire centralized governance of your business terms, context, and everything, we federate it back to Fabric. So if your agents or humans are in Fabric, they will be able to understand the context in and beyond surrounding Fabric for your agents and humans to be successful. Yaron, feel free to add anything I've missed or missed.

Razi Shafiq

30:30 Right. One other question we have here in the Q&A. Will Atlan also integrate with Fabric in such a way that we can request access to Fabric assets through Atlan?

Chandru Sugunan

30:48 Yeah. No, I, I don't I don't have— we don't have this supported today. But then the general direction is the source, the sources, in this case Fabric, or for Atlan, any other sources, could be Snowflake or, you know, wherever it is. The access control the systems have to and are building. So I don't see a reason why Atlan have to invest in that. The right place to do access control is the source system, and that's, that's where the policies, the, you know the policies and the enforcement are running. So we would prefer, you know Fabric taking it over.

31:35 And Yaron, would you do you have a different response? I, I guess my point is Atlan is not investing in any access control. We will hand it over to Fabric.

Yaron Canari

31:46 Yeah. I agree.

Razi Shafiq

31:55 Right. And the last question here that we have, I think is going to be better spent getting into a deeper session. But I'll ask it to see if, if Yaron or Chandru you might, we might have a high-level answer today. What is the difference between Fabric ontology and Atlan? Both are successful in context.

Chandru Sugunan

32:21 Sorry, difference between what and Atlan?

Razi Shafiq

32:23 Fabric ontology and Atlan.

Chandru Sugunan

32:29 Again, it is along the similar lines I mentioned earlier, right? Your ontologies... Well, I would like maybe Yaron take the ontologies, what's the roadmap for ontologies, how we are thinking about it in Fabric, and I will add on to it.

Yaron Canari

32:43 Yeah. I think that, that the question is quite the same like we, we had before. So the Fabric solution is, is focused on Fabric data, like OneLake data. So if you have a connector in Fabric and you bring the data to OneLake this is where you will get even if you don't get it today, you will get in, in through the time, you will get more. So if today we are at item-level lineage, we're investing more. But we are focused on OneLake, and this holds also for talking about ontology. So ontology gives you the context for AI agents but it's scoped to only Fabric data.

33:22 So whatever you bring to Fabric, you can also associate with the ontology, and this is the value we bring to the agents. So if your data is in Fabric, I guess most of the things and, and even the, like, I'm not dealing with the comparison between functionality that we have today because it's also changing frequently. But if your data is in Fabric, you will get a lot from the Fabric ontology, and it will, I guess, suffice most of your needs. And if not, there is a long road map for that. And where Atlan goes is that we, you want to extend that and of course, Chandru, you can take it from here.

33:58 But anyway, Atlan is giving you more than just the Fabric data estate or OneLake data estate. It can give you also the view for things that are outside of Fabric. And this is where you can bring, I guess, one ontology or one context for all of your data estate, whether it's Fabric or non-Fabric.

Chandru Sugunan

34:18 Yeah. That's, that's, that's the right answer. Yeah, yeah, thank you for, Yeah. -sharing it.

Razi Shafiq

34:25 We have two, two questions left. One I think this is for Yaron. Let me just read it. How does Atlan and Microsoft Purview work together as they are both for offering similar catalog and lineage features within Fabric?

Yaron Canari

34:50 So the question is to me about Purview and Atlan? So I think— Sorry, I think it's sorry.

Chandru Sugunan

35:00 I would take that answer. Okay. So I would say at this point I don't know what the Purview Fabric integration is. That's partly let, let's Yaron take. But then from a Atlan standpoint, we our deeper integration is with Fabric. So we have some level of platform primitives. Like in Atlan, you can bring your own assets, and that bring your own assets can be bring your own assets can be, you know, wherever it is. It doesn't really matter. Purview or your desktop or, you know, your different system on-prem, you can bring, bring that own.

35:45 But how the Purview Fabric work together is probably a question to you, Yaron.

Yaron Canari

35:53 Yeah. So I think the, the way we think about the OneLake catalog and Fabric is all like we need to... First, we need to be, bring the governance closer to the data. This is where we need to make sure that users with Fabric license should get native governance capabilities within Fabric. So if your data is in Fabric, we should gi-give you metadata, context, and and governance altogether. And, We also stick with the approach of being open, like I showed before in the deck. So we need to make sure that customers, any customer can work with the tools they want with the data they have today. We don't enforce anyone to move the data to OneLake.

36:36 We, we do encourage that. We do give additional value for those customers that bring the data to OneLake. But we are working together with Purview, with Atlan, with Collibra, with Informatica, with any, any partner or directly with customers to make sure that they can get the metadata, get the audit logs, and, and get the context so they can get to the right solution for them. So we are, we are not sticking with first-party only applications like Purview. We really insist to be open, open standards to everyone.

Chandru Sugunan

37:18 Yeah.

Razi Shafiq

37:21 One other question that There Yeah, go ahead.

Chandru Sugunan

37:23 Yeah, there are a few more questions coming in.

Razi Shafiq

37:26 Yeah. One other question that we see in the chat here, "Can Atlan be a replacement for Microsoft Purview Unified Catalog?"

Chandru Sugunan

37:34 Yeah. I... Anton, I would like to dive a little bit deeper into into, you know, your, the whole setup of what Microsoft Purview is. Would you mind dropping a note like you know, by email or whatever that is the best way to reach out to you? But of course, my, my, my answer from Atlan side is yes, but it could be a replacement for Microsoft Purview Unified Catalog, but would like to understand how your current setup is so we can serve you better with the actual situation.

Razi Shafiq

38:20 And we will finish off with "Was the lineage shown earlier constructed manually or automated specifically native Atlan feature or agent?"

Chandru Sugunan

38:31 Yeah. Yeah. We, we are deep into the weeds of anything that's lineage. We have everything, first of all, is constructed automated, constructed automatically as Razi mentioned. But we go deep into the anything that in the industry you can think of lineage. Everything that is provided out of box, like we have tiering we have, we have tierings of lineage. Anything that's provided out of box by the source system, we extract. And then we have you know, several years of investment into our query parsing technology that go mines the queries and parse and construct lineage.

39:17 And we also are the biggest consumers and contributors to the open lineage community. And we constantly both contribute to the community as well as be the thought leaders there to have you know, several of our partners integrate open lineage in their stack. So the lineage is built in a very modular and constructed way with full and end-to-end automation. So it is all fully automated. And we go very deep into lineage. I think there is another question on OneLake catalog and Purview. "What's the difference between OneLake catalog and Purview?" I think Yaron touched briefly earlier.

Yaron Canari

40:04 Yeah.

Chandru Sugunan

40:04 Do you want to reemphasize that, please?

Yaron Canari

40:09 Yeah, I think the OneLake and it also comes to Atlan and OneLake catalog, because OneLake catalog is part of the Fabric platform, and it's focused on, on OneLake data. And, and Purview data governance, I believe the question was on, on the data, data governance side, is like think, think about it like it's the catalog of catalogs. So it gets the Fabric OneLake, but it also gets data from other sources as well. So it's like the, the catalog of catalogs the way we see it. So Fabric is like a, a subset of, of those of these data sources. So they have multiple connectors like other catalogs, and, and they get data from Fabric.

40:51 And, and this is also where the difference is.

Chandru Sugunan

40:54 Yeah. And there is also another question, "Is there a two-way sync between OneLake catalog and Atlan?" The answer is yes. I think one of the earlier questions I typed in the answers. The way you can think about the two-way sync is this is, this is with Atlan, with Fabric, and all the other ecosystems. We construct the context right from the technical metadata, lineage, business metadata, and whatnot, and we surface the context to customers and the partners via via our Iceberg or Atlan Lakehouse. So we construct all the context, build our Atlan Lakehouse, which is in Iceberg format, and that is available to bring your own compute.

41:41 And that compute could be running in Fabric or Snowflake or wherever it is. And yes, that's how we sync it back to OneLake catalog. Hope that explanation makes sense, Hamza. Otherwise, I'm happy to spend a dedicated time to show and tell what this what this integration looks like. And in fact, this is one of the integrations we laun— we launched for Fabric conference, Fabcon, last week, and we are continuing to deeply work with the Microsoft team to enhance it even more with full catalog federation in the coming months.

Razi Shafiq

42:22 All right. With that said, we are winding down. If questions do continue to come up, please feel free to reach out to either us here at Atlan or Microsoft Fabric with Yaron here. But thank you for attending this Atlan in Action webinar with Atlan plus Fabric.

Chandru Sugunan

42:49 Thank you all.

Razi Shafiq

42:51 Thank you.

Yaron Canari

42:52 Thank you.