Pulumi Rocks AI-Infused Infrastructure as Code Platform
Pulumi, the Infrastructure as Code (IaC) specialist, recently introduced a bold new product, probably the most innovative addition since the company’s core product itself.
The product, Pulumi Insights, uses advanced search, analytics and artificial intelligence to bring intelligent infrastructure as code to every engineer, the company said.
Moreover, Pulumi Insights lets engineers ask any question about their infrastructure across more than 100 clouds, using either structured search queries or natural language prompts, said Joe Duffy, founder and CEO of Pulumi. Supported clouds include public clouds like AWS, Microsoft Azure and Google Cloud; cloud native technologies like Kubernetes, Helm and VMware; and SaaS infrastructure like Snowflake, Cloudflare and MongoDB.
AI Infusion
The biggest new thing with Insights is the infusion of AI throughout the Pulumi platform, such that it understands usage patterns and can deliver recommendations or even generate infrastructure as code automatically.
The company’s new Automation API embeds infrastructure as code into larger software programs and goes beyond simply generating content, the company said. Meanwhile, Pulumi also introduced a new companion website and command-line tool that taps Large Language Models (LLMs) to create infrastructure as code for any architecture for any cloud in any language, Duffy said. Moreover, red hot generative AI tools such as GitHub CoPilot and OpenAI ChatGPT already support Pulumi, he said.
“Incorporating ChatGPT into Pulumi’s infrastructure as code sets the bar for how generative AI will drive innovation in the IaC and cloud management marketplaces,” said Jason Bloomberg, an analyst at Intellyx. “Rest assured, however, that Pulumi’s competitors aren’t far behind. Generative AI is changing the game for the entire industry. Only time will tell how differentiated Pulumi’s offering will be once the dust settles a bit.”
Yet, while the dust is settling real users are using Pulumi Insights for real projects.
“Docugami develops a proprietary Business Document Foundation Model, an LLM for Generative AI applied to business-owned documents, for the commercial insurance, finance and professional services segments. We do this by operating AI infrastructure at massive scale across numerous cloud technologies,” said Jean Paoli, co-founder and CEO of Docugami, in a statement. “Pulumi’s unique automation capabilities make the task of managing all of our infrastructure seamless, and we welcome the new features provided by Pulumi Insights to help us manage our cloud even more effectively.”
Head Start
Indeed, time will tell but Pulumi has a head start.
“The cool thing about Insights is no matter how you’re doing infrastructure as code — with Pulumi, with CloudFormation or TerraForm, we can still give added value,” Duffy said. “And I think that’s a tremendous positioning as the number of developers we can now help is vastly increased.”
Pulumi Insights provides analytics and search across cloud infrastructure, generates infrastructure as code from natural language prompts, and even enables new AI-driven infrastructure automation, the company said.
“It gives folks insights into what’s happening in their organization with respect to cloud usage and cloud infrastructure,” Duffy told The New Stack. “So that includes, you know, usage, trends, costs, security — really just understanding what is happening within the organization when it comes to shipping cloud software and cloud infrastructure.”
The end result is tremendous gains in engineering productivity, the company said.
“As companies rely on more cloud resources across an increasing number of infrastructure and managed service providers, it becomes all the more difficult to locate, manage, and track resources across organizations, teams, and projects,” said KellyAnn Fitzpatrick, an analyst at RedMonk. “Often, this results in the infrastructure layer working as a bottleneck that increases the lead time from ideation to delivery. Pulumi aims to address these issues by providing tools and processes designed to enable companies to analyze infrastructure as code resources and apply that knowledge to cost control, forecasting, security, and compliance.”
Full Scenario
Torsten Volk, an analyst at Enterprise Management Associates, agreed.
“Pulumi Insights could be truly disruptive as it translates the developer’s or cloud engineer’s intent into actual infrastructure, based on business factors such as cost, compliance, security, performance, etc.,” he said.
Supporting his assertion, Volk laid out a full scenario of what is potentially possible with Pulumi Insights.
“Ultimately, you could ask Pulumi AI to build an app environment for a certain business purpose and get a recommendation and the corresponding code based on your project context, your organization’s context, and the overall industry context,” Volk explained to The New Stack. “Or you can simply ask for a new environment to run a NoSQL database of a certain size and with certain performance and scalability requirements, and the platform could make recommendations based on the different characteristics of NoSQL datastores on AWS, Azure, GCP, etc.
“When you’ve made your decision, you can ask for the cloud-specific infrastructure code to add the deployment to your git repository. At the same time, you could audit all of your infrastructure code based on certain cloud-independent criteria and receive back a report that lists all relevant deployments across clouds. Based on this report you can then ask for the code to remedy critical issues, control cost, or simply make your environments more consistent.”
Same Band, Brand New Album
Finally, Pulumi Insights builds upon the company’s flagship infrastructure as code technology to effectively become the second major release of the platform, Duffy said.
“This is our second product and the first time the band named Pulumi has an album other than Pulumi,” he said. “It’s our first time really going outside of the infrastructure as code wheelhouse, so I would say it’s the biggest launch since we initially came out from my perspective.”
