Databricks Apps for Everyone: Build Custom Apps Without Writing a Line of Code

Summary

Databricks Apps has surpassed 5,000 customers and 100,000 monthly active users, with growth accelerating in 2026 driven by the rise of AI agents.
Genie App Builder is a new AI-powered, no-code experience that lets anyone—including non-technical users—create applications on the Databricks Data and AI platform through a plain-language conversational workflow.
App Spaces separate resource access from workspace access, enabling governance at scale with cost controls such as scale-to-zero compute.

Databricks Apps for Everyone: Build Custom Apps Without Writing a Line of Code

You shouldn't need to be a software engineer or experienced data practitioner to turn your data and AI into something your team can actually use. this video previews a new native AI-powered development experience on Databricks - one where anyone can go from idea to working application through a simple, conversational workflow. We'll show how a purpose-built agent, deeply integrated with the Databricks platform, makes building apps faster and more accessible than anything a general-purpose tool can offer. Whether you're a data analyst, a business user, or a developer who just wants to move faster — the future of app building on the Lakehouse is for you.
Talk By: Fabian Jakobs, Sr Staff Software Engineer, Databricks ;

Chapters

FAQs

What is Genie App Builder on Databricks?

Genie App Builder is a new AI-powered, no-code development experience on the Databricks Data and AI platform that lets anyone build applications through a conversational workflow. It is deeply integrated with the platform, allowing users to go from idea to working app without writing a line of code.

How many customers use Databricks Apps?

Databricks Apps has more than 5,000 customers and over 100,000 monthly active users as of 2026. Growth is accelerating, driven in part by the rise of AI agents and the broader shift toward everyone building with AI.

What are App Spaces in Databricks?

App Spaces separate resource access from workspace access, so teams can share compute resources without granting full workspace permissions. They support governance at scale and include scale-to-zero capabilities to help control costs.

Who can build apps with Genie App Builder?

Genie App Builder is designed for everyone—data analysts, business users, and developers alike. Unlike general-purpose tools, it is purpose-built and deeply integrated with the Databricks Data and AI platform for faster, more accessible app creation.

Full transcript

[00:08] today talk about datrix apps for everyone um aka genie app builder. Um before I start maybe just like a small story about myself like I've been working on IDE and developer tools for
[00:21] probably like the last 15 years. I was always like excited of like building tools that allows other people to build things. Um, but up up until now these were always always like developers. Um,
[00:33] and today I'm like I haven't been ex as excited about a launch uh uh as now in a long time because now it's not just developers like everybody everybody can write apps. Everybody can build stuff.
[00:46] Um, let's get started with apps. Apps. Dick apps has been incredibly incredibly successful. We have more than 5,000 customers and more than 100,000 users
[00:59] use the apps every month. Sorry,
[01:12] accelerating. And you see like there's even is accelerating and this of of course like due to agents and earlier this year we had an offsite and like five of us were sitting down from the apps team and we were making predictions
[01:24] what is like the annual revenue by the end of the year. We are like halfway through the year. We've already blown past the medium predictions and we're very close to the highest most optimistic predictions we made. So it's
[01:36] it's it has it's been like a hell of a ride. Um, but before we get into like the serious stuff, I want to be like do something a little bit more fun. Um, so we can get
[01:48] something uh fun started. Um, I have like three questions for you. Number one, favorite color. Shout it out.
[02:02] >> Okay. Uh, pink. I take pink. Favorite animal? Monkey. >> Monkey. Pink. Monkey. Uh, favorite aart game. >> A card game? >> Uno.
[02:14] >> Uno. No, like not not card game. Ar arcade game. Sorry. >> Oh, Space Invaders. So, uh, Pink Monkey Space Invaders, right? So, let's let's just do this for a sec. Build
[02:28] Space Invaders. Uh, uh, space ship is pink. Aliens are monkeys.
[02:43] Oh, pink. Okay, let's just leave this running. Maybe we can go back to this later. There's there's some risk that this doesn't work, but uh where's the fun in having no risk? So,
[02:56] pink monkey space invaders. All right. Uh serious. Um today, our main builders are data practitioners. So these are like data scientists, data engineers, AI, ML engineers, software
[03:10] engineers and databri apps has been built with these audience in mind. Um but this is not what the reality in 2026 is. 2026 is everybody is writing code
[03:24] and everybody is building things using AI. So, so we ask ourselves the question but what what if anyone even nontechnical users can build deploy data and AI apps
[03:36] and data bricks? What would we have to change in the platform? What would have to change um for daticks apps and data bricks? Um maybe another question. Who of you has built a data bricks app?
[03:50] Okay. Now a question for those who didn't raise the hand. Who of you wants to build data bricks apps? What's stopping you? All right, let's see what's uh if we can move remove these blockers out of the
[04:03] way. Um this basically recap from the keynote. I'm going to really quick, but it's really like we have three things that are really really important. Number one is if everybody's building apps, not
[04:16] everybody is going to be an admin in the data workspace. That's not going to work. So we need to have strong governance. we need to separate uh the managing permissions, managing access to resources from building apps. Um so this
[04:29] is really important because otherwise nobody would be allowed to do those even though they would want to. Um the next one is if everybody's building apps, we probably have a lot of apps and most of
[04:41] these apps are actually not going to be used every day. Um and it would be silly if like people would not be allowed to build apps because people are like people are uh scared that they are just too expensive. Um so that is the next
[04:54] issue and then the next one is if everybody need wants to build we want everybody to build apps installing an IDE installing uh CLI code uh downloading code from GitHub is going to
[05:07] be a barrier for a lot of people. So we need to evolve the platform to to account for this and uh we've announced three solutions for these three problems. For the
[05:19] governance we introduce app spaces which allows us to separate uh access to resources from access to apps. Um so administrators can get create an app can uh grant permissions to tables to data
[05:33] warehouses to genie uh uh to genie spaces and and all these kind of things and then they can decide who can build apps using these permissions. So I can create one for my organization or for my team um for our PM team for example and
[05:48] the PM team has access to certain data and they can create apps and I can separate the creation of the app from governing access to the resources. super important. The next thing is basically no-brainer. If the app isn't used, why
[06:01] do we why do we run them in the first place? So, we introduce scale to zero. The app isn't used, it will scale to zero. It will not cost anything. Um, if I come back, I want to use the app, they start up in seconds. And then the last
[06:14] thing, which is basically what I'm talking about the rest of the talk, Genie app builder. So how can you build an app um without having to install stuff without having to manage a local machine and all that kind of stuff.
[06:32] So this is introducing Genie app builder. Genie app builder it's first party AI tool to build app for citizen developers but really for everybody. Everybody can build apps. So we have the app builder. The app builder uses uh deploys two serverless microVMs. These
[06:46] are like the scale to zero kind of apps and they use the app space where they where access to resources governed. So the builder knows what has the app access to, what can it use, how can it use them.
[07:00] Okay, now we can go to the exciting part. Let's let's do the demo. Um but before I do this, I I need to set the stage. Um so now we are not data bricks anymore.
[07:12] I'm not a data bricks engineer anymore. I work at Wonder Bricks Incorporated. Wonder is a company that uh rents um holiday vac homes like Airbnb. It also
[07:25] coincides to be like a very nice uh artificial data set on data bricks workspaces um where where we can say see like what are good destinations like a lot of really interesting rich data to build apps on. And then we have like two
[07:37] personas. We have Alice. Ellis is CEO of Wonderbricks and she's really interested in like how is the business doing. Then we have Bob. Bob is our data scientist and the data scientist basically responsible for preparing the data so
[07:51] that the CEO or the executives can actually make use of it. All right, let's uh jump to act number one, the app.
[08:08] I needed to jump over this. Sorry. >>
[08:21] like analyzing data, looking at data, making uh generating intelligence out of data. And my my starting point is not okay, I need to create this app, this this KPI dashboard for for Ellis. Um I'm
[08:35] going to start here and look what the data has to offer. And what I did here is I I I started with the genie code a regular genie code and asked it like can you give me like some interesting ways to visualize
[08:48] to visualize the data in the data uh wonder data set and came up with a lot of interesting things. It like looked at queries and created tables and joined them and at the end of the day it made
[09:00] some uh recommendations. I'm not going to go through and read this, but basically what I told it here is expose these three experiences and create a text file with instructions for an agent
[09:12] to build the app. Focus on what to build and not so much on how and add references to tables and then at the end I said like write it to a file. So this is basically this is my prompt. This is
[09:24] like the description of the app that I want to build with you together in this session. Um, again, I'm not going to read this. It's going to be very boring. But what I'm going to do instead, wait,
[09:36] is I'm going into the the builder or into the apps page. So, this is the apps page. You might have seen uh before like we you go to DRI apps here, you get
[09:48] here, but what you haven't seen before is appspaces um which is where I govern access. So, I have like my dice demo here. and I have access to a SQL warehouse to
[10:00] a G&D space uh model serving endpoint and a lot of other stuff and then in the build section I can build it. So what I'm going to do now is I'm going to
[10:13] jump to the instructions which I downloaded and I will drop it here and I say build this and I
[10:26] pick build. So you saw like I I did like all the research, all the analysis uh in Genieode uh exported the the what to build and uh I'm now going into Genie
[10:38] code uh uh Genie app builder. The first thing is it's creating an app development. So this environment this is like a microVM where all of the development is going to happen. It's not the actual app. It's like a separate
[10:50] from the app. So you can develop without actually changing the live app app because you don't want this. Um the builder is running. You see the prompt on the left and it's now basically
[11:03] reading the instructions. So it's reading the file and in the meantime the actual app um is going to start up. And the actual app that's that's starting up is like we we
[11:15] have um a library called appkit which where we encode like a lot of best practices on how to build data apps. But you really don't have to know about this. The agent has to know about this. So the agent has like a lot of instructions how to build apps using
[11:27] appkit. And what's happening now is like we start with like a scaffolded app like an starter app um which should show up any any second.
[11:39] Um, and the starter app will basically be the start for all of the apps that we built in the in the app builder. While it's doing this, we can take a look what what it's doing. Oh, yeah. No, here it's
[11:51] coming, right? So, this is the starter app. This is basically it knows that we have a warehouse. So, there's like an analytics uh page uh pre-generated. We know that
[12:04] the app space has a genie space. So, there's like a genie space chat there. I I configured model serving. Um, so these are the the starter things, but what the the agent is now doing, it's throwing it all away and replacing it with the app
[12:17] that we want to build. And uh I'll just show like a few things that it's doing. So like it's reading some instructions. It's running it's running commands. It's reading skills because we have different
[12:29] skills for for working with data bricks apps. Um it will read like data bricks. So it will like uh read the tables, get descriptions of the tables uh and at at the end of the day it will create like SQL queries a front end and uh like a
[12:43] generic back end. Um that's going to take a while but you see okay this is interesting. So so like now I have a clear picture. So it actually did like some some queries
[12:55] uh the number of bookings we have payments we have page view we have clickstream data we have destinations countries properties reviews um and now it's creating like a bunch of to-dos to
[13:07] build the app. Um what I'm going to do now I going to do like a small time lapse um jump to here. So this is actually what what it created for me when I when I ran
[13:20] this uh for the first time without any uh any addition from my side. So this is the app that that came out of this the prompt. Um we got like a landing page. We see like destinations, properties, bookings. We have destination
[13:34] performance. So we can see which of the destinations perform good. San Francisco reasonably well. Mayor York, Paris, Phuket
[13:46] a lot better. I don't know why. Um I can drill in by country. Let's say let's say go let's go to Germany. Um Germany is doing poorly. We only have Munich and Berlin. So that's not all
[13:58] that interesting. Um I can I can go into risk uh risk leakage where we we see like how what are the the cancellation rates for for c certain destinations and then traffic lab is like the click
[14:10] stream. Where does the traffic come from? Does it come from a tablet, a mobile, a desktop? And you see by this perfect distribution that this is generated data. So, but uh what I think
[14:23] is like s super impressive like even for me building this thing is I can get a really really detailed really thorough dashboard with just a text file. I really didn't do much more. Um
[14:38] okay, so I'm Bob. I'm giving this to to Ellis and Ellis is thrilled. She's like, you're doing a great job, Bob. Um, but let's go back. Let's go back into Oh,
[14:50] here let's do a quick recap. So, what what we saw is like one is Genieode is really good at understanding your data and understanding um kind of the data and AI space. So, I use Genie code for
[15:02] what it's really good at. Um, and then I use Genie App Builder as kind of a handoff from one tool to the other because the app builder is really great at building building apps, building building software. Um, and Genie app
[15:14] builder is really the easiest way to build apps. Um, I can also now go ahead and I can deploy this and I have like a running app working on my data. Really can't can't get really easier than this.
[15:26] Um, but real life, we're not you're not done yet. And this is like, "Yeah, this is a great app, but it kind of looks very average, very bland." Um, I mean,
[15:39] we we have a a company uh branding policy, but and this doesn't look anything like this. And if if you create a dashboard, there's only so much you can do in dashboard. You can tweak some colors, but uh we have an app. So, what
[15:51] let's see what we can do if you have an app. So, one thing I did is I deployed the app. So this is actually the app running here and I could now go into the app
[16:05] builder and prompt it to make it look nice. Um but it happens that our company just acquired a license for like clot design. Clot design is a very nice tool to like create like visual designs and experiment with visual designs. And what
[16:17] I do instead is like I'm going going to take a screenshot of each of these pages.
[16:30] Right? So take screenshot of all of them. And what I do then is I go to clot design and I just drop all the screenshots here. And then um let me make it a little bit bigger. Uh I have a problem that says I have an
[16:43] operation dashboard for my company wonder. Wonder bricks is travel website like Airbnb. I want to redesign the dashboard to feature a bright summer vacation style look and feel instead of the blend generic UI that we have now. Um, I'm going to start this, but again,
[16:57] watching uh claw design doing its work is kind of boring. So, I'm going to do like a time lapse again and just jump to the result. And this looks better, right? This looks a lot better. So, we
[17:10] we have this design. So, this looks nice. Risk leakage. How does this look? Yeah, the the charts look nice. Something here. Oh, this this looks beautiful. So yeah, let's what what how
[17:24] do we get this from one tool to the other? Um what I'm going to do here is um
[17:36] why do I want to reload? Wait a second. Right. Um let's go here. Okay. So what I can do is like there's a send to feature. It's called send to cloud code, but our tool is close enough to cloud code. And I can do download zip file.
[17:49] And I can download the zip file. And what it's doing is there's like a import the the attach wonder bricks read the readme inside and then implement it. Um not going to do this. I just go back
[18:02] to my to my prompt here. I go back to the zip file that I just downloaded. And uh
[18:14] how can I get it here?
[18:26] Just uh full screen is nice but uh let's do this pull it in implement this design. So all I need to do is I take the zip drag it
[18:38] into the builder implement the design. Um and now let's see what happens. So nothing happens. Oh no it happens. So it basically looks at the file. It's like,
[18:50] "Oh, this is a zip file. Let me unzip this file." So it's reading the file and and it sees there's an image. So it's reading the image that that's in there. Um it copies the the image into the
[19:03] right place. Um and it's basically going now through page by page. Um okay, it finds some issues. No big deal. Typically, if it finds issues, it will
[19:15] just recover. it will figure out itself how to uh how to get out of them uh the problems file is large let me read in chunks and view the screenshot to understand the
[19:27] design um again I could I could wait here and we could read this but what I'm going to do instead is time lapse I love time lapse and this is what it came up with again
[19:40] like uh I didn't do like any additional prompting uh if you compare it like this was the design from cloud design. Let's let's go risk leakage. Um,
[19:55] no destination performance. I like destination performance. For example, look at these gradients here and these these these things. It is pretty amazing. I I I'm always like amaze myself if I see this. But basically,
[20:07] again, I I all I did is like very very straightforward. I tell it what to do. I don't tell it how to do it. And it's just doing it. Um, right. So, let's go back into making it look good. What did what what
[20:20] did we learn? Um, wait, this the same slide.
[20:33] open tool that that interacts with other tools. Uh you can you can can use um other tools to do the design. You can import them. You can drag and drop files. You can this morning I like the for the opening slide I I took a
[20:46] screenshot and uploaded the screenshot and told it like to implement the HTML it would just do it. Um so it's it's really really easy to use and really easy to integrate with other tools. Um also there's no excuses for dashboards
[20:58] to be bland. Like there's really no excuse for like a dashboard to look boring anymore. Um okay so we built the app. The app works. uh we made it look good.
[21:12] Um but if if we a dashboard without AI is not a dashboard in 2026, right? So we have to make it smart and there are different ways to make to make something smart. Um for example, we could actually
[21:25] create a custom agent and that is something that you can totally do. Um the appkit library has like built-in functionality to build custom agents. Uh but in our case like we we know the data set. Um and what we can do is we
[21:37] basically can uh just tell it to use the genie space that we have. We have the genie space on the wonder bricks data set. So maybe just add a chat window for the genie space data set. Um
[21:51] this is the the ready app. So let's let's actually do that. Are we here? Okay. Let me create create a new chat. Start fresh. Um, I want to
[22:06] be able to ask questions about the data set. Add a side panel
[22:20] uh for walking to my genie space. Um, make it
[22:33] to level from button header. So I want to want to be able to talk to the code. I wanted to add a
[22:46] genie space chat window as a as a as a panel on the right hand side. I want to be able to talk about from the header. Um
[22:58] so almost too easy to to be true but it is actually that easy. And the first thing you see and this is like super important like uh it listed data bricks resources and what what this is doing is basically it's querying the app space that this app is deployed in uh like what resource do you
[23:12] have access to and it actually discovered oh wait there's this genie space available that I can use. So it's it knows now like it knows the genie space it knows how to access it and it also knows that the app when it's running can query the genie space. Uh,
[23:26] next thing it's doing, this is also super interesting. It's reading the AppKit documentation about how to add a Genie space to an AppKit application, which means it's saving a whole lot of tokens figuring out how to do something
[23:39] because we already told it to do it. And then there's al also like there's a genie chat component apparently. So, it's using the Genie chat component to to build the chat into the application.
[23:51] Um, it also figured out like if the the the folks of you who have built an app know that you define the resource you want to access in the app in an app.yaml and it figured out the genie space is not in
[24:03] this app.yaml. So it actually adds this to the app.yamel. And we also know that if you change that file, we have to reload the app. So it can actually re uh uh read the environment variables to to
[24:15] get to the GD uh space. But again, as a user of this app, I don't have to know this. So, it's refreshed the app. And it's going to do like a bunch of uh client changes. Um,
[24:28] let me see if I wait a second. So, it's wiring up. So, any moment this app will probably break because it's changing the code of the the front end.
[24:40] So, just see if it breaks.
[25:05] Okay. Oh, we see the button aski. It's it's it's it's not not quite pretty, but it says like, okay, now now I need to add actually CSS to make it pretty. So, let's see if it uh it can do that.
[25:18] Oh, this looks better.
[25:30] Good job, Genie Code Builder. Uh, so it updates the to-dos. It's like, okay, I'm I'm my job is halfway done. I added a button. I'm good. Um, but now the button needs to do something and it's actually actually needs to add uh the side panel. Um, I'm going to switch. Oh, well that's
[25:46] okay. So, it it basically figured out that it needs to restart the server. So, a couple of things that are that are important here that you that you notice is the agent actually controls the app. So, the agent uh can start the app, it
[25:58] can stop the app, it can reload the app. It can read any of the error messages of the app. It can actually see the app um by taking a screenshot. And I'll I should see this here. Let's take a
[26:10] snapshot. Uh, which basic Look at this. Isn't that awesome? So, it actually clicked now. So, it it it it took a snapshot and then evaluated script. It's like it's running a little bit of JavaScript inside of the app to see if
[26:22] it actually works. Um, so so what what's what's super interesting here is the the the agent will run until everything works and it will actually make sure that it works by
[26:34] actually using the app the way the end user would use it. Um, I haven't haven't seen it creating an app that's broken. It's it's really fascinating. Um,
[26:46] so now it's basically going to test a bunch of things, but I think it's it's going to be going to be good. So, what I want to do now is actually cut over another time and show you the app. So,
[26:58] this now this is the app that we have. It's uh we made it functional. We made it look good. We made it smart. So I could ask the AI now. So let's say
[27:10] what are the top five estimations by revenue and this will now talk to my genie
[27:22] space. Um and it knows about the data set. Um it it basically knows uh everything about the wonders data set and should be able to answer this question for me.
[27:45] It's not. Well, I will refresh. I'm impatient here.
[27:57] Yeah, this is looking better. Oh, 52 is the genie end point down. Okay, this is uh only half cool. Let's go back. Oh,
[28:10] yeah, this one is still reloading. Let's see if we can There we are. Let's actually try it here. Let's try it here. But if the end point is down, then this is not going to
[28:22] help us. Okay, this one works. Filtering the filtering for asking AI.
[28:46] Okay. Top five destinations by revenue. Puket, Gold Coast, Mayoka, Paris, Abu Dhabi. We see the SQL query. So, you can copy and paste this and add this to your uh your dashboard. We see a table. Okay. A chart. We can let's let's look at the
[28:59] table. We see the numbers. This is pretty amazing. And I all I did is like ask it to add a genie genie chat. What's really nice is I can just do exactly the same thing and creating a custom agent. And I can also ask him to create a
[29:12] custom agent. Uh and it will do the same thing. So this is really cool. I'm happy about this. All right, let's go back. Um
[29:27] Oh, okay. So I'll mix up the slides, right? Um, so now Alice is really, really happy. She is so happy. She has an app that uh shows her the data that looks great, that is smart and intelligent and use AI. She can ask any questions, even the
[29:40] things that are not in the in the designs. Um, and and like Bob is super happy because her boss is is happy. He gets a raise, of course. Uh but as things go like
[29:54] Ellis is not always happy with the with the app. There's always something that's nagging and maybe it's small thing and typically she would have to go to IT department or the the engineering department or the data department and create a ticket and
[30:07] and poke the people and they tell that that they're going to prioritize it and like a week later nothing happened and all she wanted to do is like move a button or something. So this is where it gets like really empowering because she
[30:20] can do it and I will I will show you how how she can do it. So let's go back to our final app. So she can do it. She can she can go into the uh the app builder.
[30:33] Let's uh let's actually show you the entry point that she would probably see. So she can go to the the Daybreaks workspace and she would uh see all the apps.
[31:06] Yeah, it's this one. She knows it's this one. And there is like a open engine app builder. So the the app is running. It's deployed. But uh you can click here and again this opens uh an app builder
[31:19] session. I need to transfer this here and and there's an app. She's like look at this like this these 500,000 page views. This looks a little bit artificial. I don't want to have it here. So she can she can go ahead like
[31:31] okay let me let me let me tell the the agent what I want. I just select this and you see that uh this will show up here. And uh I don't want this. Remove
[31:44] it. And again, I'm I'm not telling how. I'm telling what, right? This is really key. Like I'm I'm telling what. And so the agent said, "Okay, I've selected the
[31:56] page views KPI card. Let me look at the component and where it's used to remove it." Uh and we made it made it so smart that if I select something here, there's a link back to the source code where it's defined and the agent can basically just like look at the source code.
[32:08] doesn't have to do any search any anything like that and just remove it. So we should see this disappear any moment.
[32:23] Okay. So it still tells me that what it is that it's going to be removed and uh which file it is. I don't really care, but Okay.
[32:47] See what it's doing here. Yeah, it's reading a few files. So, it's found exactly where it needs to be making modifications. Okay. Oh, so it it actually figured that the
[32:59] the CSS also needs to change because there is no five column thing. So it removed it now. So now it should actually figure out that this design is actually not great
[33:11] because this is strictly worse than it was before.
[33:30] I want to wait for this. This is This is not a great way to end, right? Okay, here we got it. Here we got it. Fixed it. So, it uh figured out that the layout
[33:43] was broken. Found a way to fix the the layout. Um Ellis can now go ahead like review this change. She can go ahead and deploy this change and gonna going to show this in a second. Um but first the
[33:57] agent is pretty thorough to making sure that everything is right. So so it will basically now make a check. It's like uh is is the code right? Um when I run the server, do I see any errors? When I run
[34:09] uh open the browser window window, do I see any errors? I might take a screenshot to see if it actually did did what it was supposed to do. Oh, now it did something else. Okay, interesting.
[34:22] Um, but uh I will do the deploy. You see like it's doing the type check, the linting. Um, I will do the deploy from another app. So, um
[34:35] um let's actually do it from here. This one where we added the genie. This is the one that we actually waited out. And deploy again is we use the agent and that what what what we do is like we we create like a if you're using git we're
[34:47] using git under the hood again like you don't really need to care about that but to create a branch for this session um we will now basically commit all the changes that we made in the session merge it into the main branch and then we actually trigger like a regular app
[35:00] deploy and once the app app is deploy deployed I can uh this is the deployed app I can actually share it with my company I can share it with people um and and basically uh do the updates and
[35:13] and show it to everybody. So this is uh act number four. So what did we do? We we we created an app from uh a description of an app. Uh we made
[35:26] it look good using an external tool that we uh imported into into our builder. Uh we made it smart by adding a genie chat panel and then we showed how how you can
[35:38] tweak it. uh very visually, right? So what did we not do? We did not look at the code. We did not uh write the code. Uh we did not have to debug anything. Uh
[35:51] and that is very it's important like the the app runs exactly in the same way it would run on data bricks app. So if it runs in the builder, it's going to run. So if if you've run it locally, you might have run into issues where like your local permissions were different
[36:03] than the app permissions or the the permissions to the data is different. Not going to happen in this case. Um so yeah so every kind of everybody can make changes. Um app builder can see
[36:15] and interact with the app as almost like a user and apps are really never broken. Um that's my recap. Um summary day basically we evolved day
[36:29] apps from something that is developer centric to something that everybody can build. uh Genie app builder um targeting preview in August is going to be our first party AI authoring tool for for apps. It's going to be key to our traded
[36:42] strategy to bring everybody into uh enable everybody to build apps. And then um the the schedule zero apps uh app spaces
[36:54] is going to be central for actually being able to govern it and to control the costs. Um if you're interested, reach out to your account teams. uh they get get you signed up uh onto the uh the waiting list. Um so you get the preview
[37:07] as soon as we open it up. Um before I close it, who remembers the three questions from the start?
[37:20] >> Okay, let's take a look. So this is live, right? This is uh no safety net. Pink monkey space invaders. Here we go. All right.
[37:32] Thank you so much,

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.