Migrate PySpark Workloads to Teradata to Fast-Track AI/ML, Minimize Costs

Wednesday, April 17 2024 | Virtual

Organizations using Databricks often struggle to build and operationalize AI/ML models at scale due to its immature workload management, lack of hybrid architecture, and inaccessibility to low-code/no-code users. 

If you’re seeking a more flexible, scalable, and cost-effective solution, you can now use Teradata’s new conversion tool, pyspark2teradataml, to seamlessly migrate your PySpark workloads to an equivalent format that works in Teradata Vantage™ with minimal additional coding. With this approach, data scientists can leverage Teradata’s truly multi-cloud architecture to develop, train, and score models while limiting data movement from tools such as Databricks. This will enable organizations to operationalize AI/ML at scale with reliability while significantly reducing overall costs, complexity, and time-intensive work. 

Join this live demo to learn more about this new module in Teradata’s Python library and see a step-by-step implementation of a workflow using a PySpark-based use case entirely on a Vantage system.  

Learn how this new feature can help you: 

  • Perform feature engineering on various use cases with minimal data movement 
  • Rapidly develop and refine model features for data at scale with preferred tools 
  • Seamlessly amplify existing investments in AI/ML tools  
  • Apply strong governance, automation, and control of machine learning products  
  • Ensure future-proof AI capabilities with Teradata’s open and connected ecosystem  
  • Use open-source, partner, and native functions to train predictive models that can be used inside or outside of Vantage 
  • Score and evaluate models at scale, then prepare for easy operationalization 


Alexander Kolovos
Alexander Kolovos

Sr. Software Architect, Product, Teradata

Alexander Kolovos is a Senior Software Architect in Teradata’s Product Engineering group. His experience extends across the Languages Ecosystem products with key contributions to the R and Python integration for Teradata Vantage™ in-nodes scripting. He also authored the Teradata Orange Book titled “R And Python Analytics With the SCRIPT Table Operator.” He is the Product Owner for a range of ClearScape Analytics™ capabilities like Open Analytics Framework, third party integration, and language client libraries.