Azure Cloud Developments: Enhancing Infrastructure as Code

Neptune is now available in beta as a conversational AI agent designed to function like an AI platform engineer. It manages the provisioning, wiring, and configuration of cloud services necessary for running containerized applications.

Neptune is language and cloud-agnostic, supporting AWS, GCP, and Azure. This tool aims to bridge the gap between a deployment agent and a true AI platform engineer by understanding code and automatically planning and provisioning the required infrastructure.

It works with AI coding assistants implementing the Model Context Protocol, allowing them to authenticate, provision cloud resources, build and deploy applications, and monitor deployment statuses. The configuration is defined in a neptune.json file, facilitating a deterministic infrastructure model that developers can review, approve, or modify.

Supported resources include storage buckets, secrets, and databases, with unique physical names for each provisioned resource, enhancing accuracy in code generation. Neptune requires applications to be containerized to automatically provision and deploy, but it can assist in generating suitable Dockerfiles for non-containerized applications.

It utilizes Kubernetes and Crossplane for managing the lifecycle of resources, ensuring proper networking, identity layers, and orchestration. When provisioning, Neptune provides authoritative metadata to prevent misconfiguration by AI tools.

While Neptune is a significant advancement in AI-driven infrastructure as code, it faces competition from platforms like Pulumi, which offers a similar AI-powered agent called Neo. Neptune is currently in beta, and interested users can join the closed program on the company's website.