I have the privilege of exchanging with Chief Data Officers around the world. I’m noting a consistent trend emerge with their data science modernization efforts. Getting rid of legacy is a bigger challenge than most anticipated.
CDO’s all know that to make real-time decisions, organizations will need access to the most up-to-date data. They understand they must democratize data (both structured and unstructured), evolve processes, restructure, challenge cultures, and re-architect technology stacks to support data team success. They get that the data should be open and accessible across different tools and systems across silos — not locked into closed formats or proprietary systems inaccessible by other technologies or teams.
They are excited about the future of Apache Spark as a unified data science platform with its open source innovation in advanced analytics, ML/ AI, NLP and optimal decisioning etc.
However, a key issue is removing these proprietary/legacy systems from day to day work streams. No one quite anticipated it being this hard to get rid of SAS. Even as they brute force their way through a SAS migration, new models are being built and are relied upon by the business. One banking CDO I spoke to said he’d been trying to get rid of SAS for over 5 years. Despite best efforts, he was upset about having to renew the SAS license again with an annual 20% fee increase.
The only way to address a SAS migration is to automate it. Wise With Data ‘s SPROCKET conversion as a service is the world’s only automated SAS to PySpark migration solution.
Why automate from SAS to PySpark? To learn more about world leading migration use cases, contact us @ firstname.lastname@example.org