Domo x Snowflake: 3 Ways to Turn Cortex AI Into Instant Dashboards

Hi, today we're going to show you how your Snowflake environment can integrate directly with your Domo environment. We're going to start with Cortex Code, also known as Coco. We'll talk about Snowflake Co-work and also Domo's AI chat functionality, and how that can integrate directly with Cortex AI.

So first, what we have here is Coco, or Cortex Code. What we've done here is actually gone through and built an agent using some of our APIs, which you would be able to do as well, and then integrated it directly into Cortex so that we have a seamless integration between Domo and the Snowflake environment. We also have the ability to connect MCP directly so you can use different versions of that. From here, you can see that once we get that agent up and running—we've called it Vessel just for the demo.

Once we have that and everything is set up, I can ask it simple questions, like, "Hey Vessel, build me data around airline safety," or "build me a dashboard or an app around airline safety." You can see here that it's going through the different steps that it needs in order to gather the data. It will generate airline safety data if there is none, or if you already have some in Snowflake, you can just pull and utilize that.

Once it creates or finds that data and makes sure it is in your Domo environment, it will move on to the next steps. As you can see, we have the table created. It is then going through and using Cortex to generate Snowflake insights based on the data that exists there. From there, we have the AI-powered insights generated. We can then connect Snowflake via Cloud Amplifier or cloud integration so that the data is seamlessly transferring from your Snowflake environment into your Domo environment.

And from there, it's going to start building out the app and the visualizations that you are expecting to see. So we have the connection; the dataset has been registered. We are now going to start building that airline safety dashboard. As you can see, a lot of these choices were decided within the agents, where you can actually set guardrails and guidelines on how you want to operate, specific visuals you want to see, or maybe even color schemes based on what the company normally does in those situations.

Once done, it will publish that app to Domo for you and provide a link. It then gives you a little output based on what we've told the agent to do once it is finished. Now that we're back on the Domo side, you can see this is the app that it has created for you. We have the dashboard portion of it—Airline Safety Intelligence. You can see all the information that's pulled from that dataset: airlines tracked, safest airlines, airlines by alliance, and so on. At the bottom, you can also see the Cortex insights, which are all generated through Cortex AI.

We've also told the agent to include a "how it was built" section, so you can see step-by-step who did what piece: the stage details, the request, and what it would traditionally take to build something like this. You can see what Coco and Domo can do together, which piece was handled by Coco itself, which piece was handled by Snowflake, and then where Domo came in to put it all together. This really helps you accelerate the data in Snowflake into your visualizations to help you answer questions. And that's the goal of all three of these options that I'm going to show you here today.

The next one is Snowflake Co-work, which is generally done directly in your Snowflake environment. If we come into here, you can see it is very similar. You can connect through MCP. You can also utilize any other agents or skills that you have created, and you can just start asking questions about your data. For example, "Do I have any NFL data in my Domo Snowflake environment?" and it will pull through and show you what that is. Then, again, you can say, "Build me an App Studio app that analyzes the last three years of stats for all NFL teams." That will go through, pull all the information for you, create everything on top of it, and then go and create the app that we're expecting to see. It is very similar to the Coco situation and goes through all the same steps. In the end, you will get the app or dashboard on the Domo side that you can then share out while still maintaining all of your governance.

The last one is going to be in Domo directly. You can see here that I'm just asking it a specific question. This is all run off of Cortex AI in the background. You can integrate it and choose that as your default AI of choice. From here, you can go through, and it says, "Here's a breakdown of the partner analytics that we have for you." With the addition of toolkits and agents powered by Cortex AI, you would be able to ask it to go create a dashboard for you based on the data that's in your Snowflake or the data that's in your Domo, or a combination of the two. After you do some manipulations, it is still all run off of Cortex AI powered by Snowflake, accelerating that Snowflake data so you can make faster decisions and actions.

‍