Build AI-driven near-real-time operational analytics with Amazon Aurora zero-ETL integration with Amazon Redshift and ThoughtSpot

Every business that analyzes their operational (or transactional) data needs to build a custom data pipeline involving several batch or streaming jobs to extract transactional data from relational databases, transform it, and load it into the data warehouse. In this post, we show how you can leverage Amazon Aurora zero-ETL integration with Amazon Redshift and ThoughtSpot for GenAI driven near real-time operational analytics. 

Solution overview

The use case for this solution is near real-time analytics on transaction data from Amazon Aurora. Zero-ETL integration is set up between Amazon Aurora and Amazon Redshift. Once transaction data is available in your Amazon Redshift cluster, you can then run AI powered search on the data to get your insights using ThoughtSpot.

The following diagram illustrates the solution architecture at a high level.

sequence showing how to use Amazon Aurora's zero ETL integration with amazon and redshift to find near real-time analytics on ThoughtSpot

Below are the required steps to implement this solution:

  1. Create zero-ETL integration under Amazon RDS.

  2. Create a connection in ThoughtSpot to Amazon Redshift.

  3. Create AI powered analytics in ThoughtSpot.

Prerequisites

Create Amazon Aurora zero-ETL integration with Amazon Redshift

To create an Amazon Aurora zero-ETL integration with Amazon Redshift with sample TICKIT database, please follow all the steps in the getting started guide until your Amazon Aurora transaction data is replicated to your Amazon Redshift serverless aurora_zeroetl database. 

Analyzing near-real-time transactional data with ThoughtSpot

Validate source data in your Amazon Redshift data warehouse

Navigate to Redshift Serverless, open query editor v2, pick the cluster/workgroup and database created from integration from the drop-down list, and run the following SQLs to see integration in action, seconds later:

select * from aurora_zeroetl.demodb.sales;
How to validate source data in your amazon redshift data warehouse

Create connections in ThoughtSpot

To create a connection from ThoughtSpot to Amazon Aurora zero-ETL tables in Amazon Redshift, complete the following steps:

1. Login to your ThoughtSpot portal.

2. Create a new data connection to Amazon Redshift. 

3. Enter the connection name and description. Click Continue.

how to connect ThoughtSpot to amazon redshift

4. Enter the following parameters:

  • Host: your Amazon Redshift serverless workgroup (refer to this blog for more detail).

  • Port: 5439

  • User: your Amazon Redshift serverless user name

  • Password: password

  • Database: aurora_zeroetl

how to enter your amazon connection details

5. Click Continue.

6. Select all tables and all columns except column name start with padb*.

Select all tables and all columns except column name start with padb*

7. Click Create Connections.

8. Click Create.

how to finalize thoughtspot and aws connection

9. Your new connection to Amazon Aurora zero-ETL tables in Amazon Redshift is created.

creating Amazon Aurora zero-ETL tables in Amazon Redshift in thoughtspot

Create AI-Powered Analytics in ThoughtSpot

To create high performance AI-powered analytics using ThoughtSpot, complete the following steps:

1. Click Create new from your Data workspace and select Custom SQL view.

Custom SQL view

2. Enter SQL statement below into the SQL Editor and click Run.

SELECT qtysold, pricepaid, saletime, eventname, venuename, venuecity, venuestate
FROM aurora_zeroetl.demodb.sales AS s
INNER JOIN aurora_zeroetl.demodb.event AS e ON e.eventid = s.eventid
INNER JOIN aurora_zeroetl.demodb.venue AS v ON v.venueid= e.venueid

3. You should see Results returned.  

returned results

4. Enter a name for the SQL View and click Save.

5. Click Exit.

6. Click Search data

7. Click Select a Data source.

8. Select the SQL View created and click Select.

Select the SQL View

9. Select saletime, venuecity, pricepaid, and qtysold and click Go.

Select saletime, venuecity, pricepaid, and qtysold and click Go

10. Click the 3 dots and select Save.

Select save

11. Click Save answer.

describe your answer

12. Click Search answers and Liveboards from the top right corner.

search answers and liveboards

13. You can now leverage natural language processing and generative AI to search insights from your transaction data from Amazon Aurora. For example, “what is the qtysold over period of time?”

Turn your insights into action

It’s simple to set up an Amazon Aurora zero-ETL integration from Amazon Aurora MySQL to Amazon Redshift. This integration eliminates complex data pipelines and enables near real-time analytics on transactional/operational data. You can then connect ThoughtSpot to Amazon Aurora zero-ETL integration tables in Amazon Redshift to run AI-Powered Analytics.

Try it for yourself! We look forward to hearing from you about your experience. If you have questions or suggestions, please leave a comment. Otherwise, see how ThoughtSpot is working with AWS to deliver AI-Powered Analytics and faster insight-to-action.