Unstructured Data in Data 360 Follow Along Project Exercise

Key Takeaways

Okay, trailhead brothers here. Uh, today we will do the branch unstructured data in data 360. So let's go right ahead. If you go to the first get uh started with unstructured data in data 360, you have to create your own playground from here. Just click and it will send you an email, reset your password and then connect it here. Okay. What it's saying is here is there is an unstructured data. Uh it can be any data. It can be audio chats, email, knowledge articles, anything. And we get that data from external uh storage and what what it does it creates a relationship between the external and the internal. What it does is it creates a unstructured date uh data lake object and it map it uh to the unstructured data model object. Okay. So in here uh what we have done is for example from external data we have referenced the data and we have created an unstructured data lake object and then we mapped it to unstructured data model object. So model object is uh in the salesforce this is from the outside what it does and that's what we have done there's a one one relationship and there can be many to one relationship. So that's what what saying it it is here. After creating the playground, you can just click on verify and this challenge will be completed. Okay. Let's go to the next step. Okay. Ingest knowledge article data. What we are going to do here? We are going to use the trail head uh trail head to uh ingest or put dummy data for us so that we can test. What we have to do? We have to go to the data cloud. Enter data. Click the data cloud app. Okay. Go to the data stream tab. Okay. So let's go and launch our playground. Now we're going to put some uh data from the outside to the inside. All right. And that's what we are doing here. We are just putting some dummy data into our Salesforce account. Okay. If you launch your new playhead trail, uh just click on the nine icons on the left side. Uh click here. Just search boot data and click here. Come here. After coming here, uh click on the data streams that we have to go here. Okay. And this see the next steps. Click new. Select the seal source data. Click next. Okay. on the top right just click new in the data streams we have to go and then click new in here just click this Salesforce CRM this is the demi data and then click next Okay. Okay. We have to select the knowledge data bundle and then we have to click next. It is loading all the data. Click uh new leave the default. Select select the knowledge data and click next. Okay. In here we have to select the knowledge data. Uh okay there it is knowledge data bundle. This is the one that we have to select knowledge. As you can see they have written here. And uh let's just uh click next. Okay, you can see all the fields that are selected here and uh review the fields. I think this is fine. Click next and click deploy. Okay, now leave it as it is. This is all the fields that are coming with it. Click next. and uh then we have to click deploy. What we are doing here we are um uh we are deploying our data stream and we're waiting for it uh to be ready when the status is active. So what it does is the knowledge article data lake objects that we have uh imported here or deploying here. What it does is uh it is converting uh it to it to unstructured data lake objects and uh in the data lake objects tab uh in the data 360 we can uh we also view the unstructured data lake objects that uh it is creating in the data lake objects tab when it is active. You can see data stream is active and uh everything status is active stream. So we have ingest the data and uh let's uh select all click on the top right to update the status and it will update all of the the status here. Okay. Data stream status is good. If as you can see in the data lake objects we will we can see our data here you see it has converted our data into on structured data lake objects. So now let's go and verify our data. In the next step, what we are going to do is we're going to create a vector search index configuration for our knowledge article object so that it is easy for the for any uh for the flow and for agent for Apex to easily retrieve and search our data. Okay, assessment is complete. Let's go to the next step. Okay, in here what we are as you can see we have data sources we ingest the data and it will convert that data to data model object and unstructured data model object and what we are going to do is we are going to chunk the data and generate vector embeddings to create vector search index so that it is easily for the Salesforce uh uh softwares and applications to search the data and get the result quickly. So that's what we are going to do do here and uh if you go come below you can see uh let's go to index and click new okay you can see here index uh I do not see let me check uh click is in the detail then search index you have to search index Let's select here. This is the one search index. Click here. Okay. Then if click advanc setup next advance. Okay. Okay. Yeah. Click new here in the search indexes object. We are going to create a vector search index. Okay. And here click on advanc setup. Click next and then advance setup. From the select source object, select vector search then article version data model object and click next. New tab will open. Uh we have to click uh vector search and then knowledge article this one data model object and then click next. Okay. Okay. Then we have to click uh next here on the top right. Okay. As you can see it is chunking. Uh let's click on uh manage fields. Click all fields and click save and then click next. Okay. Let's select our fields to chunk. Let's select the article number. Save. Uh passage instruction just leave it as it is. Uh click next. Okay. Select all you save and uh click next on the leave the default vectorzation strategy and click next. Click knowledge article version DMO. Uh let's go back. I think uh this is the one knowledge article version. Uh this is what they are saying knowledge article version. We have to select the knowledge article version and then click next. In here click manage fields. Select all the fields and click save. Okay. Leave it as it is and let's go to the next. Okay. And in the vector. Okay. Okay, let's leave it as it is. The vectorzation strategy. Uh, click next. Search related fields for search filtering and okay, just click next. Okay, now it is ready. Our chunky materialization we have defined everything here. That's going to replace the autogenerated search index configuration name with my cave. Okay. Search index uh conversion name. Okay. Let's replace this my Okay. And then we have to click save. Let's click save. Okay, we have created the index data model objects. Now we will test it. We have to make sure that the search index status is ready. Uh let's go to data explorer data model object. Okay. After uh loading uh we have come here to our search index. Our search index is created for us as you can see. And uh then we can go to data model object in here as you can see our uh data model object for our chunk and index has been created for us. So this is what it is. Now we can use our data for gen AI purposes uh anywhere. So it is easy for the Salesforce applications to get our data. Now just verify our step. We can finally find Minecraft chunk object. This is submitted to in progress to ready. Let's go to search index in here. As you can see our search index is in progress. Uh okay. when then if it is ready then we are good to go. So we have to wait for a couple of minutes. Okay, search index is ready. Uh verify our challenge. Congratulations. Challenge is completed. Thank you.