Organisations are racing to implement AI, with 88% of enterprise leaders expecting to increase spending on generative AI over the next year. However, siloed IT teams are often a barrier to successful AI delivery.
Data science and AI teams have long operated as shadow IT, explains Rhys Powell, senior black belt, Managed Cloud Services at Red Hat, speaking in a CIO webcast.
“Infrastructure teams kind of ignored them in the past,” he says. But as AI adoption has become such a key strategic priority, “now is the ideal opportunity to pull them in, support them, align them with your application development teams.”
This means AI teams should be integrated into mainstream DevOps workflows.
“What you can do is enable both parts of the organisation to work together on the same infrastructure, utilising the same methods,” Powell says.
That means using the same continuous integration and continuous delivery/deployment pipelines, GitOps practices and observability tools that application teams have used for years.
This increases flexibility because “you’re able to land the right workloads in the right location” and means “security teams are going to suddenly see a lot more adherence to the rules that they require.” “You can automate your processes, you can automate your tools,” Powell says.
An opinionated platform like Red Hat’s OpenShift, which allows organisations to smoothly deploy new AI applications while managing the underlying infrastructure for you, makes it easier for IT teams to collaborate and deliver effectively.
“It abstracts away the complexity so that you can focus on the value creation that you need to do to make your customers happy,” Powell says.
Click below to watch the full webcast series:
