In Forrester’s recent report, DataOps for the Intelligent Edge of Business, Michele Goetz, et al., describe how data teams are facing challenges “about data to support the return on investment and experience.” In fact, “no amount of investment in new big data systems, cloud migration, modern data warehousing, or data integration will completely solve the problem. The approach to data is shifting toward DataOps.”
We agree and you can read more about why DataOps matters in our blog, For Data Team Success, What You Do is Less Important Than How You Do it.
Below we provide additional suggestions for further reading based on Forrester’s principles for advancing DataOps.
Prioritize the quality and value of deliverables – “To ensure data products succeed, adopt a test-driven development protocol to create tests upfront and maintain repeatable unit tests that can also be markers for upstream policy compliance.”
Speed up delivery for shorter development cycles – “Agile development strategy shifts the goal post for deliverables from complete platform solutions to smaller products defined by quality and value-based milestones.”
- Blog: Minimizing Analytic Cycle Time with DataOps
- Case Study: Accelerating Analytic Cycle Time with DataOps
- On-Demand Webinar: Orchestrate Your Development Pipelines for Fast and Fearless Deployment
Build for reuse, flexibility, and elasticity – Data “products become building blocks for a variety of analytics and application solutions…”
Govern data by design – “DataOps addresses data governance policies through the creation of rule-based services and processes.”
- Blog: Continuous Governance with DataGovOps
- Blog: Govern Self-Service Analytics Without Stifling Innovation
- White Paper: Governance as Code
- On-Demand Webinar: Redefining Data Governance with DataGovOps
Executive through inclusive teams – “DataOps works in synchronous and asynchronous fashion with DevOps, ModelOps, and data governance teams.”
- Blog: Improving Teamwork in Data Analytics with DataOps
- White Paper: Reducing Organizational Complexity with DataOps
- White Paper: Warring Tribes into Winning Teams: Improving Teamwork with DataOps
- On-Demand Webinar: Your Model is Not an Island: Operationalizing Machine Learning at Scale with ModelOps
Forrester concludes with recommendations for technology investment, in particular for lineage, impact, and root cause analysis. “Vendors such as DataKitchen are addressing this problem with detailed views of data flows and error rates.”
For more information, you can read the complete Forrester report here.