Taboola calls itself a "discovery platform" or a "reverse search engine." It delivers personalized content recommendations to more than a billion users who visit many of the web's most popular websites every month. You've probably seen Taboola content yourself – a block of images and headlines labeled "Sponsored Links by Taboola." The links are designed to increase user engagement with the sites on which they appear.

Panoply provides end-to-end data management-as-a-service. It uses machine learning and natural language processing (NLP) to learn, model, and automate standard data management activities from source to analysis. It can import data with no schema, no modeling, and no configuration. Users can quickly spin up an Amazon Redshift instance and run analysis, SQL, and visualization tools just as they would on a Redshift data warehouse they created on their own.

Developers can pull data out of Taboola's servers using its Backstage API. For instance, you could fetch information about a campaign by calling GET /backstage/api/1.0/[account-id]/campaigns/[campaign-id]/items/.

The response from Taboola's Backstage API comes in JSON format, and might look like this:

{
  "results":[
      {
      "id": "1",
      "campaign_id": "124",
      "type": "ITEM",
      "url": "http://news.example.com/article.htm",
      "thumbnail_url": "http://cdn.example.com/image.jpg",
      "title": "Demo Article",
      "approval_state": "APPROVED",
      "is_active": true,
      "status": "RUNNING"
    }
    ]
  }

When you've identified all of the columns you want to insert, use the Reshift CREATE TABLE statement to create a table in your data warehouse to receive all the data.

Once you have a table built, it may seem like the easiest way to replicate your data (especially if there isn't much of it) is to build INSERT statements to add data to your Redshift table row by row. If you have any experience with SQL, this probably will be your first inclination. Think again! Redshift isn't optimized for inserting data one row at a time. If you have a high volume of data to be inserted, you should load the data into Amazon S3 and then use the COPY command to load it into Redshift.

At this point you've coded up a script or written a program to get the data you want and successfully moved it into your data warehouse. But how will you load new or updated data? It's not a good idea to replicate all of your data each time you have updated records. That process would be painfully slow and resource-intensive.

Instead, identify key fields that your script can use to bookmark its progression through the data and use to pick up where it left off as it looks for updated data. Auto-incrementing fields such as updated_at or created_at work best for this. When you've built in this functionality, you can set up your script as a cron job or continuous loop to get new data as it appears in Taboola.

And remember, as with any code, once you write it, you have to maintain it. If Taboola modifies its API, or the API sends a field with a datatype your code doesn't recognize, you may have to modify the script. If your users want slightly different information, you definitely will have to.

Panoply is great, but sometimes you need to optimize for different things when you're choosing a data warehouse. Some folks choose to go with Amazon Redshift, Google BigQuery, PostgreSQL, Snowflake, or Microsoft Azure Synapse Analytics, which are RDBMSes that use similar SQL syntax. Others choose a data lake, like Amazon S3 or Delta Lake on Databricks. If you're interested in seeing the relevant steps for loading data into one of these platforms, check out To Redshift, To BigQuery, To Postgres, To Snowflake, To Azure SQL Data Warehouse, To S3, and To Delta Lake.