Extract Structured JSON Data from Unstructured Data using Databricks Mosaic AI

Back Next

Step 4. Publish and run the guide

You can enter field names based on the context of an uploaded file to extract data.

Open the Extract Fields with Context
guide.

On the

Start

tab of the Start step, ensure that the

Run As

field is set to

Current User

Save and publish the guide.

On the Actions
menu, click Run
. Alternatively, you can copy the execution URL from the Properties Details
dialog box to run the guide.

On the

Instructions

page, select the context file that you want to upload from your machine.

In the Enter field names to extract data
field, enter comma separated field names to extract data based on the file contents.

Click

Continue

You will receive a response from the LLM.

You can extract new data with the same context or upload a new file to start a new conversation.

To extract new data, click

Next Question

. Enter new field names with the same context.

To upload a new file, click New File
. Upload a new file from your machine and enter field names based on the file contents.

Click Continue
.

You will receive a response from the LLM.

If the uploaded file is empty or the field names are not related to the file context, an error occurs.

You can also use the embed code to embed the guide into an HTML document of a third-party application.