GenAI calls for more rigorous data discipline than previous forms of analytics, but there are lessons platform teams can learn from existing approaches.
DataOps early adopters were inspired by DevOps principles to help data scientists quickly create business value from big data; as generative AI apps go mainstream, that interplay between IT disciplines is coming full circle.
As with Agile software development and DevOps methods, DataOps sought to break down organizational barriers and encourage collaboration between business stakeholders and IT teams. Thus, there has been some exchange of techniques between disciplines already. DevOps and platform engineers have already applied methods developed by DataOps and machine learning operations (MLOps) pros to AIOps and observability workflows for years.
But now, generative AI (GenAI) is pushing data-driven analytics further into the mainstream among business users, bringing data management and data governance to the forefront of enterprise IT ops concerns.
Adoption of AI, including GenAI, was cited as the top driver for increased usage of corporate data in recent market research. Of 318 respondents to a June 2024 survey by Informa TechTarget's Enterprise Strategy Group, 56% rated AI as the main reason for more users of corporate data within their organizations. AI adoption also prompts more intense scrutiny on governance and security, concerns at the heart of DataOps principles: 36% of respondents added new data governance roles and expanded existing roles over the last year because of AI.
As early adopters of AIOps and AI automation tools have already discovered, data quality and integrity can make or break such initiatives -- 60% of survey respondents classified these as "very high" or "high" priority for GenAI projects. A majority of survey respondents, 51%, indicated they don't yet fully trust or somewhat distrust the accuracy of the data used in decision making.