A leading U.S. mortgage-finance enterprise supports millions of homeowners and renters nationwide. The organization maintains market liquidity and stability by partnering with lenders across the country to make homeownership more accessible. Focused on data-driven innovation, it continues to modernize its technology ecosystem to improve efficiency, reduce costs, and accelerate insights across its operations. 

The Challenge

The company relied on a Talend-based data platform that had become costly and complex to manage. 

  • Paid $1.7 million annually for Talend licenses, with renewals projected at $3.6 million 
  • Processed more than 200 nightly valuation and analytics jobs that delayed insights 
  • Ran all jobs on a single EMR cluster without orchestration 
  • Struggled to scale efficiently as infrastructure expenses rose 

The objective 

Migrate from Talend to AWS PySpark to cut costs, boost performance, and improve scalability. 

The solution  

Accelerated migration and performance optimization using EZConvertETL and AWS automation. Wavicle collaborated with the client’s data-engineering teams to co-design and deliver the migration from Talend to the PySpark framework on AWS. 

  • Automated 80 percent of job conversions with EZConvertETL, generating YAML files for PySpark execution in AWS EMR 
  • Migrated 218 ETL jobs and 51 process launchers within six months 
  • Tuned Spark configurations to enable parallel processing and improve throughput 
  • Built a Lambda-based Resource Manager that calculates the number of EMR clusters required, scales up for peak workloads, and shuts down idle clusters 
  • Enabled dynamic orchestration by analyzing job type, runtime, and load patterns 
  • Reduced idle compute time with auto-shutdown after 30 minutes of inactivity 
  • Delivered the solution within AWS GovCloud to meet strict compliance and security standards 
  • Worked side by side with the client’s engineering team to enhance Spark performance 

The Result

Modernization improved speed, reduced cost, and strengthened scalability. 

  • Saved $1.7 million annually in licensing and infrastructure costs 
  • Improved orchestration and parallel execution across clusters 
  • Delivered the project 50 percent faster, six months instead of twelve 
  • Increased performance and simplified maintenance
  • Modernized and validated more than 200 ETL jobs 

This four-month engagement showed that enterprise ETL migrations can be fast, accurate, and collaborative, even with tight controls and changing scope

Related Posts

  • Microsoft Fabric
  • Microsoft SQL Server

Greenhouse Grower Modernizes Data and Insights ...

  • Amazon Elastic Container Service (ECS)
  • AWS Aurora

Travel Center Operator Accelerates Access to Da...