Really excited to announce Pulumi Insights today. Pulumi Insights provides intelligence, search, and analytics over any infrastructure, in any cloud across your organization. • 🧠 Intelligence: New AI-powered tool for learning, discovering and building cloud infrastructure with Pulumi, using the power of generative AI and Large Language Models (LLMs). Check out pulumi.com/ai. • 🔎 Search: Multi-cloud resource search, powered by the Pulumi resource supergraph, which surfaces rich metadata and relationships between cloud infrastructure. Available now in the Pulumi Cloud at app.pulumi.com. • 📊 Analytics: Visualize Pulumi resource data with your favorite BI tools, including Excel, Tableau and others, using your own data warehouse for reporting on cost, compliance, and other operational use cases. Read more about Pulumi Insights at https://pulumi.com/blog/pulumi-insights. A big Thank You to all the folks here who have been part of the preview programs for these features, and have helped shaped the work on these products. 🙌

are these AIs trained on our code?

are these AIs trained on our code?

No. They are currently leveraging pre-trained OpenAI models, along with in-context learning to tune them to the task of generating useful Pulumi IaC code and maintaining memory so that code can be edited.

What happens to that memory? Is it incorporated into the model or is it erased?

What happens to that memory? Is it incorporated into the model or is it erased?

It is used only during your browser/CLI session.

Thank you for answering my questions. 🙂

@late-petabyte-3251 very reasonable questions, it's something that should be quite clear in the docs IMO

Those screenshots were an example of a dashboard in a BI tool (Metabase in the example, but could be Tableau, Looker, PowerBI or even just Excel), connected to a Data Warehouse (Snowflake in that example, but could be Redshift, etc.), which had several data sources stored in it which could be joined to produce useful analytics that go beyond what the Pulumi Cloud offers directly: 1. Pulumi resource data (from the Data Export feature of Pulumi Cloud) 2. Cost data (from a cloud pricing data provider, either pricing APIs from your cloud provider, or another cost management service) 3. Departmental data (mapping of Pulumi Teams to organizational units in your business) The Pulumi Cloud does expose all the resource data, but doesn't itself (yet) have it's own pricing data. For today, you can bring your own pricing data and then join your Pulumi resource data with that.

Luke is going to do an AMA on Pulumi Insights next Wednesday (5/3) at 9am PDT. Come and get your questions answered on Insights or Pulumi in general . https://www.reddit.com/r/pulumi/comments/130b4rn/ama_with_luke_hoban_cto_on_pulumi_insightsai_at/

What's that in GMT/UTC for those not in the US?

16:00 UTC

We are going to get started in 27 minutes https://www.reddit.com/r/pulumi/comments/136qlrh/ama_luke_hoban_cto_on_pulumi_insights/