On-Demand Webinar · 1 hr 6 min

Elevate Your Data Strategy with DataOps

A data strategy is about more than the next new tool. James Lupton, CTO and founder of Cynozure, joins Chris Bergh to work through where DataOps fits in a strategy, what its key components are, and practical examples of it in use. Recorded December 2020; updated August 2026.

Presented by Chris Bergh

What you'll learn 7 points
  • Cynozure defines a data strategy as a framework that lets an organization generate business value from data and analytics with pace and agility, and builds it on six pillars: vision and value, operating model, people and culture, technology and architecture, data governance, and roadmap. DataOps belongs inside that strategy rather than beside it.
  • The obstacles named are almost all organizational rather than technical: silos of people and silos of systems, poor collaboration inside the data team and between the data team and its customers, slow processes, inflexible architectures, technical debt, bureaucracy and politics, and access to data.
  • The team-effort slide sets a data team without DataOps at 97 percent of effort on data and analytics development against 3 percent on data operations, and a team with DataOps at 85 percent against 15 percent. The argument is that the smaller development share produces more, and that software engineering already runs a higher operations ratio than data does.
  • DataOps engineering builds five things: meta orchestration across tools from one view, automated testing in production and development, environment management, better sharing across roles and teams, and measurement of quality, deploys, tests, errors, and SLAs.
  • DevOps and workflow tools alone do not add up to DataOps. Beyond continuous integration and deployment, the list adds continuous self-service environments, continuous meta orchestration, and continuous testing and monitoring.
  • Adoption runs in six steps rather than a big bang: educate on the ideas, find a first project, establish a community of interest, demonstrate real value in a month or two, iterate onto more use cases, and expand into a staffed center of excellence with common tooling and metrics.
  • Success is measured in two areas: production metrics covering build and data provider errors, test result history, timings, SLAs, and model metrics; and team and project productivity metrics covering collaboration, deployment frequency, and test coverage.

Slides

32 slides

Questions from this session

What is a data strategy?

Cynozure's definition is a framework that enables an organization to generate business value from data and analytics with pace and agility. Their model rests on six pillars: vision and value, operating model, people and culture, technology and architecture, data governance, and roadmap. The strategy is the thing that turns those pillars into a sequence of moves rather than a wish list.

How does DataOps fit into a data strategy?

DataOps sits inside the strategy as its delivery method, combining modern software engineering practice, agile delivery, and value-focused cross-functional teams. Its value shows up in three places a strategy already cares about: risk reduction, efficiency and cost saving, and driving value. Adding a DataOps engineering function is the concrete step.

How do you start DataOps without a big bang?

Six steps. Educate the organization on the ideas, find a first project that can demonstrate value, establish a community of interest and align with data and agile leaders, demonstrate real value in a project of a month or two, iterate onto more use cases, then expand into a staffed center of excellence with common infrastructure and metrics. Each step is meant to earn the next one.

Is DataOps just DevOps plus a workflow tool?

No. Continuous integration and deployment are only part of it. The full list adds continuous self-service environments, continuous meta orchestration across a diverse toolchain, and continuous testing and monitoring of data as well as code. DevOps and workflow tools on their own will not produce DataOps outcomes.

How do you measure a DataOps transformation?

Measure two areas. Production metrics cover the pulse of the current build, data provider error and success rates, test result history, timings, SLAs, and machine learning model metrics. Team and project productivity metrics cover collaboration, deployment frequency, and test coverage. The four focus areas those roll up into are cycle time, error-free days, team collaboration, and process measurement.

Where to go next