Showing results for 
Search instead for 
Did you mean: 
Employee Alumni
Employee Alumni


An AI/ML platform can generate scored data and the result of a prediction in a production system.  This data can then be shared within Incorta for creating analytics dashboard and reports.

For this article, we will use the integration between Incorta and TAZI as an example to show how the data can be shared and loaded to Incorta.

Use Case

Customer Churn:  Incorta hosts the customer data including the customer profile, customer interaction with the deploying company, customer order history, customer billing and payment history, and even enriched credit data sourced from outside systems.  The data is shared with TAZI and the ML that was built within TAZI.  In an AI/ML platform like TAZI, the prediction can be generated after the ML inference  job is executed.  The result of predicting if a customer may churn with estimated likelihood, can then be used to enriched the original customer data as well as being used in filtering and sorting customers.  The corrective action can also be triggered from this additional insight.



Incorta and TAZI are installed together within the same network.  The disk level sharing is available.

Configure TAZI to save the result

1.  Go to the business models, Go to CHANGE IO 



2. Add Destination





Select parquet, and give the path:


Use the Incorta DataLake Connector to load the data

Go to Incorta

3. Add Data Lake Data Source 

Go to Data  -> + New -> Add Data Source->Data Lake - Local Filessuxinji_3-1652206789622.png



Enter the same path as the Data Source


4. Create a schema

Go to Schema -> + New -> Create Schema


5. Create a table

Click on +New -> Table -> Data Lake


Select the Data Source you just created. Select File Type: Parquet. Enable Wildcard Union. Give ​​Include: *.parquet




6. Create Dashboards

Now you can create dashboards in Incorta. 

Go to Content  -> + New -> Add Dashboard -> Add Insight





Select the data sets you just created. 





Best Practices Index
Best Practices

Just here to browse knowledge? This might help!

Version history
Last update:
‎06-01-2022 09:09 AM
Updated by: