Quickstart (UI)

​

Overview

In this guide, we’ll develop an API to identify the top mentioned features in sales calls.

We’ll process sales call transcripts to extract insights like competitors mentioned, features mentioned, and whether it was successful. Our dataset, a 70-row CSV file, captures these activities.

Out steps include:

• Creating a workspace
• Uploading a CSV file
• Creating AI columns
• Filtering, aggregating, and refining the data into a top 2 list of features
• Exposing the top features through an API

​

Create a Workspace

We start by creating a new workspace.

If this your first time using Airfold, you will be directed to our welcome page where you can simply type in a name:

Otherwise, you can create a new workspace by clicking on the top-left corner, then

”+ New Workspace” :

​

Uploading a CSV file

We will be using sales call transcripts for this guide, which you can download here

Click on “Integrations” on the left column, then select “File upload”:

From here, upload the sales_calls.csv file we just downloaded.

Then, name it sales_calls_csv and click on “Create Integration”.

Within a few seconds, all of the data should be ingested:

​

Creating AI columns

Now, we can start gathering insights from our data with AI columns.

Start by creating a Table by clicking on “Tables”, then selecting sales_calls_csv:

We can call this sales_calls_analysis.

From here, we can start creating AI columns:

”+ Add column” > Select template: ”+ Blank column” > “Next”

Let’s create a column to indicate whether our call was successful (we define a successful call as one that scheduled a follow-up call).

Fill in the boxes like so:

You can copy and paste the description here:

Whether this call was successful (which is defined as successfully setting up a follow up call). If the follow up scheduling is successful, the call should include an explicit date/time for the call.

Column name is the name of your column

Column type is the data type of your column (we use “Checkbox” in this case since success can be a true or false value)

Description is a detailed description of what you wish to extract

Default value is what the value defaults to

Select related columns are the columns that are relevant to extracting this data

Now, click on “Create” > hover over the column name >

Your table should look like this:

Repeat this process for a few more AI columns:

To see which competitors are mentioned in the call:

Description:

Who are the competitors mentioned (if any)?

To see which features are mentioned in the call:

Description:

Which features are mentioned in the call

Select options:

'Customizable Templates', 'Inventory Management', 'Marketing Automation', 'Customer Support Integration', 'Social Media Integration', 'None'

Your table should now look like this:

​

Query Top Features

Using our AI columns, we can perform further analysis to see which top 2 features were mentioned in successful calls.

We will be using successful and features_mentioned to do this.

Click on “Queries” and create a new query get_top_features:

From here, let’s create a new node:

Copy the SQL query here:

WITH
	arrayJoin (features_mentioned) AS feature
SELECT
	count() AS total_successful_rows,
	topK (2) (feature) AS top_features
FROM
	sales_calls_analysis
WHERE
	successful = TRUE

Click on “Run”, then we can get the top 2 features mentioned in successful calls:

​

Publish as API endpoint

To publish this as an API endpoint, simply click:

“Publish” > “View Endpoint”

From here, you can use your endpoint in JavaScript, Python, or cUrl by simply copying and pasting.

You can also unpublish your endpoint anytime by clicking “Unpublish”.

​

Next Steps

You’ve successfully uploaded data, created AI columns, and extracted insights using Airfold in a few simple steps! This workflow enables intuitive interaction with unstructured data, transforming raw call transcripts into actionable insights.

Check out how to do this through the CLI through Quickstart (CLI).

Explore workspaces, sources, and queries to learn more about how to interact with Airfold. Our the API and CLI documentation is also available for further information.