Kubernetes management platform vendor D2iQ has become the latest vendor to stir generative AI into its product.
The update to the D2iQ Kubernetes Platform (DKP) is the AI Navigator, an assistant that the company says is aimed at closing the skills gap faced by enterprises on the way to cloud technology adoption.
Making Kubernetes even a little more straightforward to deal with is a laudable goal, even if engineers used to the salaries commanded by the technology might blanch at the arrival of generative AI. In the case of DKP 2.6, the AI Navigator is effectively a natural language interface for the company’s knowledgebase.
For example, a user can ask a question: “I lost my kubeconfig file to my DKP cluster. How do I recover it?” The AI Navigator will swiftly respond with a solution and often the commands that need to be executed.
Behind the scenes, the service is using Microsoft’s Azure OpenAI service and the ChatGPT 3.5 model, “though we are still exploring other options,” Dan Ciruli, VP of product management at D2iQ, told The Register.
The AI Navigator has also only been trained on D2iQ’s internal knowledgebase, although plans are afoot to make the suggestions more specific. Ciruli told us: “We have a planned enhancement to add the ability for the AI Navigator to use the customer’s own contextual data (including the configuration of the cluster, workloads, etc.) in order to provide very specific analysis and recommendations.”
Noting privacy concerns, Ciruli added: “We are exploring different ways to do this that could allow privacy-sensitive customers to get these benefits, including running the model on their own infrastructure.”
D2iQ is the latest example of a vendor looking at ways that generative AI could ease a company’s transition to the cloud or cloud infrastructure.
Noting the growth and diversity of the Kubernetes environment and containerized applications, Deepak Goel, CTO of D2iQ, commented: “Issues that are relatively minor with a single cluster become much more difficult to manage in a multi-cluster, multi-cloud environment.”
Tools such as AI Navigator go some way to addressing challenges faced by organizations seeking to adopt cloud-native technology. While they are currently very much at the level of an enthusiastic assistant armed with a search engine and a knowledgebase, the D2iQ roadmap indicates an intent to take things further.
Other cloud infrastructure companies, such as Hashicorp, are also making use of AI services. Adam FitzGerald, vice president of developer relations, said in 2022: “With more narrow domains like cloud architecture definitions and DevOps, I expect it to be possible for artificial intelligence services to learn and recommend best practices and common architectures.”
As 2023 has progressed, FitzGerald’s predictions are more prescient than ever. ®