What is Databricks Apps?

Summary

Databricks Apps is the fastest way to build, deploy, and govern production-grade applications directly on the Databricks Data and AI platform, turning data and AI/ML models into interactive, custom applications for the entire enterprise.
Three primary use-case patterns for Databricks Apps are context (surfacing data insights), agentic assistance (AI-guided workflows), and action (business process automation).
The platform supports open frameworks and coding agents for developers, the Genie App Builder for no-code creation, and on-behalf-of-user authorization and App Spaces for enterprise governance.

What is Databricks Apps?

Databricks Apps is the fastest way to build, deploy, and govern production-grade applications directly on the Databricks Platform. In this foundational session, we’ll explore how Databricks Apps allows you to transform your data and AI/ML models into interactive, custom applications that your entire enterprise can use. Whether you are building AI-powered agents, real-time dashboards, or internal tools for business process automation, discover how we’ve simplified application development on Databricks.
Talk By: Cong Xu, Product Manager For Databricks Apps, Databricks ;

Chapters

FAQs

What is Databricks Apps and what is it used for?

Databricks Apps is a capability on the Databricks Data and AI platform that lets teams build, host, and govern interactive, production-grade applications directly on their data and AI/ML models. Use cases range from AI-powered agents and real-time dashboards to internal tools for business process automation.

How does Databricks Apps integrate with Unity Catalog?

Databricks Apps is Unity Catalog-native, meaning every app automatically benefits from the platform's governance model for data access and permissions. This ensures that applications built on the lakehouse inherit the same security and compliance controls as the underlying data.

What are the three use-case patterns for Databricks Apps?

The three patterns are context (surfacing data and insights for users), agentic assistance (AI-guided workflows that help users take action), and action (automating business processes end-to-end). These patterns cover a wide range of enterprise application needs on the Databricks Data and AI platform.

What hosting options does Databricks Apps provide?

Databricks Apps runs on managed serverless infrastructure, including serverless micro apps that scale automatically. This removes the need for teams to manage containers or configure compute, simplifying deployment to production.

Full transcript

[00:08] Yeah, nice to meet you all. We're not going to do intros in person for everybody, but um thanks for taking the time. This is introduction to databicks apps, and my name is Song. I know it looks like it should be pronounced Kong,
[00:20] but it is song, trust me. Um and I'm a product manager with database apps. And so we'll chat a bit about apps today. But before I get started, I was curious to get a sense for the
[00:33] room. So how many of you have basically just heard of data bricks apps and and never really looked at the product or have no idea what it is? Should be the majority of you. Okay, good, good, good, good, good. Um, so you're in the right
[00:46] session. That's good. I'll go through sort of what database apps is what we're trying to do and how many of you have sort of tried to dabble a bit maybe have built one app and are couple
[00:58] how many of you built more than one app because you should not be you should be in a different session
[01:10] going to try to keep this well not interactive for the first half we'll do 20 minutes where I'm talking at you just to get some information across us and then we'll do a live demo that as live demos are may or may not work but at
[01:22] least it's live and then we'll do some Q&A. Does that sound good? Thank you. Thank you. Awesome. So, as with every good presentation, we're going to start with a bit of math. Okay?
[01:38] But it's pretty simple math. So, 1 plus 1 equals 3. I really think this represents apps quite well. And I'll get into why in a moment. But basically with database apps, by the way, the timer
[01:50] hasn't started, FYI. Um, but I can look at my own watch. Yeah, basically with database apps, we take two things and we make a hole that's bigger than the sum of its parts. So
[02:02] what are those two things? What is oneplus 1 and then what is three? So let's talk about it. The first thing that I'm going to talk about briefly is going to shock you all. It's a new tool called AI. Some of you might have heard
[02:15] about this, but basically in a nutshell, AI has democratized creation, right? To some degree. And the chart on the right, you might have seen on the on your right, your right, you might have seen
[02:29] this is GitHub commits. And so last year, GitHub had about a billion commits. This year, based on some projections, not all, the expected number of commits ranges from 14 up to
[02:41] 50 plus billion. So the message here is clear. People are creating, they're building with AI. However, as we all know in this room, we've probably played around played around with with AI and coding agents. It is very good at
[02:53] building common artifacts, right? Things that are out there in the world that is in that training data set. It can build quite well. I'm sure you've all also experienced something more niche or more unique that it doesn't build quite well. So AI democratize creation for common
[03:08] artifacts. That's part one. Part two is a proven and powerful platform and I will let you guess what the name of that platform is. You should have all heard
[03:20] of it. Data bricks, right? If you have not heard of that, you're definitely in the wrong meeting. But so what is this powerful improvement platform? Really, it is a bunch of infrastructure. Ali talked about it this morning, right? A
[03:32] lot of infrastructure that's really for the agent first era. On top of that, you have your data, your AI. On top of that, you have a bunch of governance, security, permissions, alles,
[03:45] things that you and your platform teams have invested a lot of time into making sure it's properly governed and secure. And so what all that confluence of things, what all that leads to is a is latent business insights is what I
[03:58] called it. There's a bunch of business insights that are inside the database platform. And they're latent because they're guarded by interfaces, right? a set of interfaces that you might know how to interact with notebooks or other
[04:11] interfaces that are somewhat rigid and not everybody can have access to those interfaces or knows how to use those interfaces. And so in a nutshell, what that means is you have this platform that's very powerful. There's a lot of insights and
[04:25] you want to democratize access to those insights, right? You would want more of your peers and your stakeholders to be able to get access to this intelligence and insights that are inside of data bricks. But what you don't want to do is
[04:38] you do not want to create all this investment into the platform that you have done. So you do not want to recreate governance. You do not want to recreate the data and AI somewhere else that you have inside of data bricks. You do not want to figure out other ways to
[04:52] get the right engine for the right workload. And so what do we do? Right? So the question is what do we do? Because on the one side again we have this new tool, this new technology that
[05:04] democratizes creation of common artifacts. Everybody can create and then on the other side we have this powerful platform with all these insights that are waiting to be democratized but it's
[05:16] not quite accessible today. Right? It's sort of obscured away behind these rigid interfaces. And what we would want in the middle is something like this, right? We'd want a common customizable interface that is
[05:30] native to the platform. So common so that AI agents can build it. Like we said, they can't build everything easily, but they can build common interfaces. Customizable so that everybody can use their preferred way to interact with
[05:44] these insights, right? It's a customizable interface to the use cases and the needs of the end user. And then native to the platform, so you don't have to recreate the platform. You don't have to move governance out. You don't have to worry about the infrastructure.
[05:57] You don't have to figure out how to, you know, get the data and AI out of data bricks into somewhere else. So I will let you guess what the answer is, what this red thing in the middle
[06:09] is. Does anybody have an idea? The answer is data bricks apps. So that's why we build data bricks apps is to connect these two forces that we're seeing. 1 plus 1 equals three, right? We
[06:23] really think there's a lot of synergy here and so ultimately apps is something we launched in 2025 at G8 last data and AI summit and the mission is to democratize data and AI through custom
[06:36] apps. So that is at the very high level what we're trying to do. Democratize data and AI through custom apps. Right? Take what you have in this platform and let deliver it to all the end users that can create with AI. By the way, we will give
[06:50] you time to ask questions at the end. I should have said that. Um but if you if you want to ask questions throughout, please feel free. Um so that is the mission. That is the high level.
[07:10] Say it again. Sorry. >> Can Can you or will you be able >> Can you or will you be able to compare this to database bricks? >> To data bricks. What? Sorry. >> Bricks. >> To data bricks. >> Agent bricks.
[07:22] >> Agent bricks. Agent bricks. Yes. I I'll I will talk about that in a bit. Yes. Maybe I'll talk about at the end. Maybe it's not natively in the presentation. >> There is it's it can they can both be used together. Yes, for sure. The
[07:36] question was can can I talk about the relationship between this and agent bricks? I will make sure to get to that uh at the end. So before we do before we dive into sort of you know what database apps is, what are some of
[07:50] its capabilities? I always think it's a bit more helpful to look at some examples, right? So what are customers building with database apps? I'll have a couple case studies and then I'll go into a couple examples at the field
[08:02] build. Obviously no customer data that I can actually show you. So first example aircraft carrier aircraft carrier operator and manufacturer
[08:14] they're building an app for inflight entertainment analysis. So you can see in the app here, they basically have an airplane with all the seats, right? And so they can look at what every person is watching on the airplane, which is very
[08:26] scary to me because I like to keep that to myself, but they know that I've watched Wicked and then Wicked for Good back toback and that's okay because they're a customer. So they're using database apps to build an inflight
[08:38] entertainment app. Another customer, very different. Bridgestone, they make tires. They have a database app for labeling images of tires and letting experts basically
[08:52] identify which images have damages and then they label them using the app and then that gets fed back into a machine learning model to train. So a very different use case. And then the third one, EasyJet, another airline
[09:06] coincidentally, they have an app to basically price flights. And you can see here they have a lot of data. And on the right hand side, they have some actions they can take. They can either increase or for the benefit of all of us
[09:18] hopefully decrease the price of a flight. So inflight entertainment price increase tool, tire labeling, very very different use cases that all leverage database apps.
[09:30] And I also want to show you a couple that are actually deployed apps. Again, not customer data obviously, but these are apps are field built. So you can actually see, you know, these are just normal apps. And you can see here in the
[09:42] URL, it's a database apps URL. This is a CPG retail insights app. So what's cool about this is there's obviously different ways you can deep dive into the data. But one thing I think is really cool, I hope this works,
[09:55] is there's a generate brief button here. So you can generate a brief and there's going to be an a agent running in the background. Watch this not work now. I hope this works. That um there we go. Generates basically executive brief um
[10:09] for the audience or for the end user. Another thing I think is cool is you have a detailed breakdown here and right here you have food storage analysis and you have it by location, right? Just a
[10:21] very neat app that looks at basically various different CPG retail capabilities. Another app here. Oh, that's the same app that I just opened. Another app here is a retail site lab, which what I think
[10:35] is cool about this one is it basically looks at where should we open our next retail site. And a cool action you can take here is you can run these demand forecasts or scenarios and you can
[10:48] trigger all these different uh you can manipulate these different categories like how far is it from urban or suburban areas and you calculate demand and you can see here some of these pop
[11:00] up as green and so that those are sort of locations that are favorable given your criteria and again this is also a deployed database app Um you can see here in the URL as well. So
[11:14] couple of examples from the field as well and apps has been a big success since we launched it. I think if you were in the keynote today, Ali talked about it briefly. I thought it was quite
[11:26] nice. You said it was the fastest growing product currently. It's definitely somewhere up there. We don't have the exact statistics versus uh some other products, but we have 5,000 customers building apps and we have about 150,000 apps. This is just a set
[11:40] of logos that you see EasyJet, you see thawless. I actually have never looked at this page in detail. Some of the logos we've talked about before, but a lot of customers are using apps and um it's growing very quickly.
[11:54] Okay. So, you've all built apps and you've all thought about apps in some capacity, I'm sure, and you've all had other solutions that also are app platforms to some
[12:07] degree. So what's special about databicks apps? Well, yeah, in other words, what makes a database app? So you saw the architecture, but this is this little thing is the apps part of the market,
[12:21] but obviously database apps doesn't stand alone. Database apps is very much embedded into the broader platform, right? So apps are a persistent custom experience,
[12:33] but they are on top of the database database platform. So this looks quite complicated, but really what it means is that everything underneath apps is something that apps can natively integrate with. And so if
[12:46] you want to simplify this a bit, the way we think about it is really apps allows you to connect to your data and AI. That's that layer that has Genie Genie ontology. It also allows you to inherit all the
[13:00] existing governance and security that you have on your platform today. So all the egress controls you've set up all the alles permissions and privileges all this very critical
[13:12] permissions and privileges that you've set up you can inherit and then at the bottom you know you have lakeflow lake base lakehouse we really think that is going to be the app back end of the future right apps today a lot of them just need a database we have that
[13:25] database lakebased but we think with agents that need more and more context apps in the future will have a lot more complicated backends they will need longunning analysis. They'll need the lakebased. They'll need real-time stuff.
[13:37] They'll need data from different places. And so the bottom is really the back end of the future. And database apps natively integrates with all of these. So that is what we think makes database apps unique.
[13:50] Yeah. Okay. I think people there's a question in the back. Yes. Just >> wondering
[14:02] >> I'm really just having people ask questions so that you know. >> Thank you. >> I'm wondering if data bricks apps is like can it be associated with a more flexible platform that you can do things
[14:14] that AIBI can't do. For example, like more sophisticated time series analyses or what if analyses. >> Yeah. So you the question is basically am I hearing you correctly? You're
[14:26] asking is apps is a good use case for apps like a more sort of niche or specialized visualization tool. And so the answer is yes. And it's actually perfect segue into I've had one person
[14:38] leave only so far. That's good. Perfect segue into the next the next part of the presentation. I have not forgotten about agent bricks. I will get to it. So again apps use cases they seem it seems
[14:52] like there's many apps use cases we had inflight entertainment we had pricing increase app we had an app that looks like tire labeling so how should you all think about what should be my first use case in database apps or my second
[15:05] sounds like most of you have not gotten beyond your second so first or second use case right so there's different use cases there's very different end business outcomes but they all often combine one of three And the first one
[15:17] is similar to what you just mentioned. Ultimately apps for us is still a human first product. So we believe that even if in the future we each have hundreds of
[15:29] agents doing work for us, right? We will need a way to control and orchestrate these agents and we are the ones ultimately responsible for the agents decisions. So we need to make sure we know what these agents are doing. And so
[15:42] the first part is apps allow you to have customized context for the human, right? We each process data a bit differently. Some people have very custom visualizations, maybe either very high
[15:54] scale visualizations. We've had that where you know AIBI was not able to meet some of the scale that some of our customers have or you know specific ways you visualize your data. Whatever it is that you need to get context about the
[16:07] ground truth, what is happening on the ground without any agentic intervention, right? You need some way to anchor yourself to know what's happening without having to rely on agents. If agents orchestrate agents and then get
[16:20] checked by agents, we're sort of in this loop where, you know, who knows what the truth is. So, we believe humans will continue to need customized context and that's one of the big use cases for apps. It can be standalone use case
[16:32] where the app is just a very niche way to visualize data. But that's not as powerful as combining that then with agentic assistance. So we're in the age of agents. And so of course we want to leverage agents to help us do our work.
[16:46] And we see agentic assistance in many forms. The most common is the chatbot. We've all seen it, right? The chatbot is a very common part of database apps where you can chat with your data. I should have shown it. One of the sales dashboards had a ask genie which is just
[16:59] a chatbot. But it also is agents doing longunning work in the background and then exposing that. That's a bit like well the executive brief wasn't that long running but it's you know you let the agent do a
[17:11] bit of work or they do work in the background they label some stuff they call out some potential areas of concern and then that gets surfaced to you and you as the human can make the decision. Also I would have never thought I'm like
[17:23] talking like differentiate between humans and agents but here we are now you know. Um, so agentic assistance is the second part that a lot of our apps have. And the third and maybe most impactful part that differentiates apps
[17:35] from everything else inside of data bricks is the ability to take action. So you derive insights from the app. You look at the data, the ground source of truth, ground source of truth, ground truth, source of truth or ground truth.
[17:48] Pick your pick your favorite. And then you let the agent help you in highlighting areas you might want to investigate deeper, maybe areas you should follow up on and that gives you some insight and then you want to take
[18:00] an action, right? So in the case of EasyJet, the action is decrease flight prices hopefully. In the case of the Bridgestone app, it would be like label this tire a certain way. So that is the third and maybe the most
[18:13] impactful piece that we've seen customers leverage apps for because it allows them to really drive business outcomes versus just staying at the analysis level.
[18:25] So those are some of the characteristics and so what categories does that translate to? Just to give you all some ideas again the universe of apps is infinite and ever expanding such as the real universe I guess but you know some
[18:39] categories and um we see a lot of these built on top of database apps and a lot of these are unique to databicks because they leverage other parts of the of the platform. So for example, real time ops
[18:52] monitoring will leverage a lot of our real-time infrastructure. Zero uh zerobus some of you heard about zerobus or coming in real time mode. Some of the
[19:04] infrastructure that we've talked about today will be leveraged for realtime ops monitoring apps. Physical ops monitoring apps one of my favorites. A lot of our customers actually I'm curious how many of you have like a real physical footprint for your business either like
[19:17] retail stores or you know like networks or whatever it is like a physical reality. Yes. Good amount. So a lot of our customers data is really the representation of that physical reality and so they use apps to monitor
[19:30] networks. So for instance, energy energy grid network energy grid network again energy grids and that is a very good example where they'll look at the energy grid in the app. They'll see some areas
[19:42] that might need further investigation and they'll have a bunch of agents also looking at the same data and additional data and making recommendations. And then they might send a field team to go check out what's going on. and that field team will use the same app to
[19:56] collect data and input it back into data bricks so that that data can then update the status of those locations. It's a really cool app. Obviously, it's a customer app so I couldn't show it, but um that's one that I, you know, it's very representative of the kind of apps
[20:09] that people are building on data bricks, but yeah, there's many different kinds that are not listed here. A common one is also business vertical SAS. So a lot of folks are extending or building new ways to do sales dashboards or to do you
[20:24] know what their marketing or what their finance SAS software used to do. So that's a very common one as well. But again the key thing is those three things context for humans
[20:36] aentic assistance and insight. So if you're thinking about hey when should I leverage database apps think about those requirements. Cool. So then let's briefly talk about what is data bricks apps like what is the
[20:49] product actually what is what are the capabilities and then I'm happy to dive deeper into this after the session as well I don't want to spend too much time on this because this is a level 100 uh oh is it a level 200 like an introduction to database apps
[21:02] so it's a platform to do really three things it's a managed platform to build to host and then to govern apps build host and govern those are the three things you want to get right most
[21:15] importantly actually govern. I don't know I didn't pull the room here but I'm assuming most of you when you think about prol proliferation of apps you're very nervous about governance and so that is also our main priority. So let's go through each three each of the three
[21:28] very quickly. Build the key two takeaways here are twofold. The first two bullets are around being open. We want it to be very easy to build using any framework. So
[21:40] any node or Python framework are supported for the app. We don't have any proprietary frameworks that you have to use. Whatever framework you want, you can use it to build the app. And we wanted to be able to build the app with any coding agent. Right. So
[21:53] >> sorry, >> do you do RS? Do we do R as well? R is like top of the feature request list for us. We know our audience. We know a lot of you love R. We actually I just had a meeting with Pit actually um to try and
[22:05] find a deeper integration with R. But the that is not because we don't want to support R. It's just a sequencing thing, but open frameworks and then any way to build. We know developers love their tools. We don't want to constrain them.
[22:17] So you can use coding agents. I'm actually going to show off using a coding agent. So that's not what we want to differentiate. That being said, we also have our own platform native solution, right? So that you can out of the box have everything. So the first
[22:30] one is database appkit. That's our own framework. It's nodebased. You can use it with your coding agents. It basically just out of the box makes the apps a bit more production grade because it knows how to properly authenticate these apps,
[22:43] properly connect to all the databick services. And then what we're announcing tomorrow, so it's a bit of a sneak peek. Well, it's not a sneak peek because Ali already talked about it today, but that's just how it goes here. Um, is the
[22:55] Genie app builder, which is a way to build apps inside of the database platform. It's a vibe coding platform, a vibe coding interface. Think of replet bolts very similar to that. The main purpose of Genie app builder is really
[23:08] to go from prototype to production, right? You can v code your interface, but then when you need to connect it to data, etc., you can finish a job if you want to use Genie App Builder. Of course, you can go end to end as well, but once again, we don't want to
[23:21] constrain what tools you might love, you know, to prototype things, etc. So, that's on the build side. Again, feel free to come up and ask me afterwards. I I'm mindful of time. So,
[23:33] on the host side, very important. It's a managed container infra. You don't have to worry about this. This is effectively serverless for you. So, what this also means is you can just go into your data bricks UI today and spin up an app,
[23:45] right? There's no longunning process. you have to go through to figure out an AKS container or whatever it might be. This is all managed by us infrastructure for any app. That means any scale, right? We have larger instance types
[23:57] now. We have horizontal scaling across these instance types. So scale should not be an issue for you to build your app on top of database apps. Stable deployments, frankly, it's a checkbox item. We don't even have to talk about
[24:10] it. And then we're also announcing tomorrow and this Ali did not give away yet serverless micro apps. And what that effect effectively is is if you have apps that are infrequently used, the
[24:23] compute will scale down to zero. It's not a long running instance. And then if you need to use it again, the app will scale back up within say 10 seconds or so. So it's not the most latency. It's not the best latency, but if you have
[24:35] infrequently used apps, you don't have to pay for the compute. So that's what we're announcing tomorrow. service micro apps and then last govern most importantly last and not just not least but last but most
[24:47] importantly governance first one managed off basically data bricks acts as the IDP you don't have to worry about that second one I really want to highlight govern with unity what
[24:59] does that mean you all have spent a lot of time making sure that all your database end users have the right permissions to the right data in the AI you can inherit all of that in the app, we have this authorization model called
[25:12] on behalf of user o. It just means that if user A and user B have different permissions inside of data bricks, their permissions carry over into the apps you're building. So user A will open the
[25:24] app and see different data than user B. You don't have to do anything to make that happen. It just takes what you already have set up inside of data bricks. That is one of our most popular features for obvious reasons and that is
[25:36] one that we think is is a real differentiator. Lastly, out of the box observability, we basically have our system tables, but we also have recently allowed for all the telemetry collected at the app to be
[25:50] flushed into Unity catalog in hotel format so that you own all your telemetry data. And then the last thing we're announcing tomorrow, maybe the biggest announcement is app spaces, which is effectively a
[26:02] way to govern a group of apps as opposed to just singular apps. Right now, a lot of this stuff happens at the singular app level. You can configure it but we know admins and central teams you want to govern groups of apps. So you can use
[26:14] app spaces to do that. Okay. Okay. So we run a bit
[26:27] Yes. Let's get to the demo. So what I'm going to do is I'm going to quickly I'm going to start the demo. Some of it will take a bit of time. So then we can already do questions and answers and
[26:39] then we'll see if you know clock cooperates or not today. We'll see. Oh, there we go. So I have a split window here unfortunately uh because we're
[26:52] using our own company internal version of cloud and so it doesn't allow for dangerously skip permissions. So I have to keep approving things. Um, but first of all, just for you to understand, this is our internal version of claw, but it's just claw code.
[27:06] And I'm going to first tell it, do you have skills related to data bricks apps? You know, this is live because I'm misspelling things and I will continue to misspell things. So, oh, it needs to run login first.
[27:19] Great. You can see this is live. Let me see if Isaac will let me log in.
[27:33] While it does this on this side, we have the app UI. And so you can create apps programmatically via cloud code. You can also go into the app UI. And the first thing I want to point out is this is not the workspace UI. I'm
[27:47] sure you all very familiar with the workspace UI. Apps is up here in the switcher. Have you all seen the app switcher? How many people have seen the app switcher? between 9 and 14. That is there's an app
[28:00] switcher. We have a bunch of other things that are not in the workspace UI. And so database apps is one of them. So when you click here, you get to this page. And these are some of the things we'll talk about tomorrow, but today you
[28:13] can create an app right here. Let me see if this worked. Do you have skills about data bricks apps? Yeah, perfect. Oh no, not perfect. Hm. I
[28:26] don't even know what's going on here. Okay, I'll show you the pre-loaded screen in a second. So, back to the UI. You can create an app very easily and very quickly via UI. And I'm just going
[28:38] to walk you through the custom way to do that. So, you can have a sense for what the things are you can configure at the app level. So, you create create custom app. You give it a name.
[28:53] You can give it a description. And then next thing you can do is you can configure a git repo. And so what you can do if you're an administrator is you can enforce that all apps in a workspace are only deployed if they have a git
[29:06] repo associated with them. So this gives you some level of control to make sure that all apps go through a CCD process that you have set up saying GitHub actions. This also supports GitLab and Azure DevOps. This is not just GitHub.
[29:20] So it gives you some control over what kind of apps get deployed. For now, we're going to skip that. Um because I don't have this enforced at my workspace level, so I can skip it. And then now are some of the things I just talked about. You can basically
[29:33] configure for your app, the kind of resources you want to attach to the app itself. So the app has access to that regardless of who uses it. For example, you can attach a SQL warehouse. You can attach a database.
[29:45] And then you also have what we call user authorization scopes. And this is basically when you want the app to take on your permissions and privileges and act on your behalf so that you can see
[29:59] all the data you have access to. You can configure here for what API scopes the app is allowed to do that. So even if you have permissions to look at certain data but you don't give the app that scope it won't be able to do it. So you
[30:12] have another lever here to control the on behalf of user authorization and then you have app telemetry. We talked about this and then you can choose your compute size enable horizontal scaling
[30:25] and after this you create the app. And basically what that will do I'm not going to configure anything right now. It's going to create some compute that we manage and it's going to you know spin up a base image that we manage
[30:37] and then you can deploy your app code on top of that right so any framework built from anywhere again meant to be open we have 9 minutes so let me try one last time to figure out what is going with my clock actually I'm just going to show
[30:50] you I don't know why I have to log in here but I ran this earlier today somewhere I think here yes so I ran this earlier I don't know if I can have a time stamp
[31:02] here. The point is all I told it is built me a database app using New York City data and it loaded a skill and by the way these skills are open source. They're right here. Database agent skills open source. It
[31:15] loaded the skills. It asked me what offer I wanted to use and then it basically built the app. It asked, you know, do you want it to be more analytical or lakebased? And then I said analytics and it built the app. And then I asked it to deploy
[31:28] it. It checked. And then I said yes, deploy. And it deployed the app. And the app's right here. Um, and this is basically the app itself.
[31:41] And it's just a simple analytics dashboard that hopefully will load in a couple seconds.
[31:54] But you can also see this is the app's details page. And so you will see here I deployed this earlier today at 7 a.m. in preparation for this. And then you can see on the authorization front it has you know a couple default scopes
[32:07] but it didn't give us any access to the user um O. You can see here it has access to one second.
[32:22] Wait, I'm trying to find one thing. So this you can see here it has access to a SQL warehouse that you can use for the app but then no other resources. You can see the instance size. So this is
[32:34] very similar to the screen that you just saw. So I don't know why my claw code refused to authenticate right now. But broadly speaking, two ways to create apps. You can use the UI. You can use any coding agent you want. And tomorrow
[32:47] there will be a third way. Genie app builder. Right. So with that said, we have seven minutes just to wrap up. Databicks apps. It's about data and business intelligence for all. You can
[32:59] build, you can host, you can govern on top of data bricks your apps. And as a reminder, it's not just an app. It's a database app. So you can leverage your data and your AI. You can inherit your
[33:12] existing governance and security from the database platform. And you can leverage the app back end of the future, which is anywhere from batch longunning jobs to millisecond zero sort of latency
[33:24] real time infrastructure. And that's it. I think yes, data and business intelligence for all.

Learn more about the Databricks Data and AI platform.

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