TLDR: Pulumi has launched Neo, its first AI agent designed to automate complex cloud infrastructure management and platform engineering tasks. This new offering promises to significantly reduce operational time, enhance compliance, and boost infrastructure delivery for DevOps teams by leveraging agentic AI capabilities.
Seattle-based cloud engineering leader Pulumi has officially launched Neo, an innovative AI agent poised to transform infrastructure management and cloud platform engineering. Announced on September 16, 2025, Neo is designed to autonomously handle intricate infrastructure tasks, aiming to condense workflows that traditionally took weeks into mere minutes, all while upholding stringent enterprise governance standards.
Pulumi Neo distinguishes itself from conventional automation tools by functioning as an ‘AI teammate’ within the Pulumi Cloud platform. It possesses the intelligence to understand complex dependencies, execute changes, continuously monitor outcomes, and ensure compliance across the entire infrastructure lifecycle. According to Pulumi CEO Joe Duffy, “things that used to take weeks can now confidently be done in minutes,” highlighting the agent’s potential for dramatic efficiency gains.
A cornerstone of Neo’s design is its robust built-in governance. The AI agent meticulously adheres to all existing team settings, including Role-Based Access Control (RBAC) and policy-as-code configurations. Users are afforded granular control over Neo’s autonomy, ranging from fully protected, human-approved steps to entirely autonomous workflows, a concept Pulumi terms ‘progressive autonomy.’ This flexibility allows teams to dictate the level of oversight for critical infrastructure modifications.
Neo is built upon Pulumi’s established infrastructure-as-code (IaC) foundation, leveraging large language models—primarily Claude—integrated with enterprise-grade guardrails. It seamlessly integrates with popular AI developer tools such as Devin, Cursor, Windsurf, and VS Code via Pulumi’s Model Context Protocol server. The agent operates across diverse cloud environments, including public, private, and hybrid setups, supporting major providers like Amazon Web Services, Google Cloud, Microsoft Azure, and Kubernetes.
Early adopters of Pulumi Neo during its beta phase have reported significant improvements. Customers have seen a 10x increase in infrastructure delivery with existing teams, a 75% acceleration in application deployment, and an average 90% reduction in policy violations. Practical applications include remediating thousands of compliance issues and upgrading dozens of Kubernetes clusters with zero downtime.
Pulumi’s commitment to this AI-driven future is evident in its development efforts; CEO Joe Duffy reallocated a third of his 130-person workforce to build Neo. The company also partnered with AWS, utilizing its Bedrock Agent Core service for the agent’s development. Sheila Gulati, Managing Director at Tola Capital and a Pulumi investor, emphasized the market need, stating, “While developer teams race to adopt AI, platform teams are drowning in managing the infrastructure demands, and a volume/velocity crisis develops… Infrastructure provisioning is still manual because of compliance priorities internally. This creates a compounding bottleneck.” Pulumi Neo directly addresses this bottleneck by automating these critical platform engineering tasks.
Also Read:
- Cloudelligent Earns AWS Generative AI Competency, Bolstering AI-Driven Cloud Solutions
- Shift Technology Introduces ‘Shift Claims’ Platform, Harnessing Agentic AI for Insurance Claims Transformation
Founded in 2017, Pulumi boasts over 3,700 customers, including industry giants like Nvidia and BMW Group, and its IaC tool sees over one million downloads weekly. With Neo, Pulumi is making a substantial bet on agentic AI to redefine cloud infrastructure automation.
