DataKitchen
DataKitchen Written by DataKitchen Authors

Behind A Great Company, There is DataOps

Behind A Great Company, There is DataOps

Recently, Amgen purchased the pharmaceutical unit of Celgene responsible Otezla for $13.4 billion. Otezla is not your typical run-of-the-mill M&A blockbuster. They are a DataOps pioneer. DataKitchen is proud to count ourselves as a key partner in Otezlas DataOps success story, which you can read more about in our Celgene case study. We have worked with the Otezla team since 2014 and are proud to be part of this multi-billion dollar success story!

Otezla proves that DataOps delivers on its promise:

  • 50+ integrated data sets in production on a daily, weekly and monthly basis

  • 1000+ data quality tests run every build

  • 12 data features added per week, on average

  • Very, very, very few errors or missed SLAs

  • Small staff supporting hundreds of sales reps, executives, and data scientists/analysts

  • Data engineering and DataOps staffing (build, test, deploy) of TWO Full Time Equivalent employees — well below the industry average

Amgen Otezla

DataOps is an analytics development and operations methodology that produces an order of magnitude improvement in analytics cycle time and quality. Below is an image of a DataOps process that creates data marts for a team of data scientists and analysts. Automated orchestration acquires data from sources, cleans it and loads it into a data lake. The data is transformed and populated into data warehouses and data marts, accessed by teams of data analysts, BI analysts and data scientists who create insights with analytics tools. When making a data architecture or other change, the related step is quickly updated and the whole orchestration simply re-executes from end to end. This automated process allows the team to quickly leverage new data sets and implement new features. DataOps controls code and data quality with automated statistical process controls, inspired by lean manufacturing. Tests check input data, as well as inputs, outputs and business logic at every step of the pipeline. With DataOps, data teams move at a higher velocity and with improved quality, to the delight of users and stakeholders.

Learn More at: