DataOps (Data Operations) is collaborative Data Management practices “focused on improving the communication, integration, and automation of data flows” to create predictable delivery and change management of data, data models, and related artifacts. DataOps were created as a better way to develop and deliver analytics, and to aid the data engineers, business uses, analysts, and data scientists within an organization. As Dan Potter notes, “DataOps is not a technology, but the right technology, people and processes to accelerate time to delivery and agility to respond.”
Other Definitions of DataOps Include:
- “A new Data Management strategy that creates a collaborative process for managing data, people, and technology in an efficient manner.” (Itamar Ankorion)
- “An emerging discipline that brings DevOps teams together with data engineer and data scientist roles to provide tools, processes, and organizational structures to support the data-focused enterprise.” (CIO)
- “A type of work focused on 12 DataOps Principles.” (DataOps Manifesto)
- “A focus on the end-to-end delivery of data.” (Forbes)
- “A rapidly emerging discipline for organizations who struggle with management of data as a shared business asset.” (Forbes)
DataOps
Use Cases Include:
- A Fortune 500 insurance provider moved to automated ETL scripts.
- A company needs support for end-to-end needs of machine learning.
- GlaxoSmithKline (GSK) Research and Development needs a better data environment to develop new pharmaceuticals.
- NerdWallet’s personalization of its content, tools and actions for their customers.
Businesses use DataOps to:
- Build out proper, agile modern Data Architecture
- Increase
data trust across the organization, leading to greater efficiency
- Improve
integration to multiple data sources
- Remove
bottlenecks between data
- Format
data so that those analyzing it can make better decisions
Image used under license from
Shutterstock.com
