Pulumi Delivery and Pulumi Agents automate post-merge workflows by letting developers and autonomous systems write delivery pipelines as code and supervise live environments. This infrastructure-as-code foundation relies on widely understood languages like Python, TypeScript, and Go to boost the success rates of automated tasks. Meanwhile, rapid adoption has already altered engineering realities, with Anthropic's daily monorepo merges jumping from 2,500 in May to over 30,000 by September. To keep these automated processes secure, the architecture relies on deterministic planning, scoped credentials, and automated policy enforcement.
Key points
Traditional post-merge operational workflows rely heavily on manual tickets and scripts t…: Traditional post-merge operational workflows rely heavily on manual tickets and scripts that create severe engineering bottlenecks.
Defining pipelines in standard programming languages enables artificial intelligence mode…: Defining pipelines in standard programming languages enables artificial intelligence models to safely deploy and manage infrastructure.
Automated delivery systems enforce safety during high-volume operations through determini…: Automated delivery systems enforce safety during high-volume operations through deterministic plans, scoped credentials, and policy checks.
Sources checked by our analysis pipeline: 1
Introducing Pulumi Delivery and Pulumi Agents: agentic code to production, and back again.
At Pulumi, we like code. A lot. And we aren’t the only ones: agents like it too. This year, the share of infrastructure operations on Pulumi done by agents went from zero in January to over 40% in October. Anything expressed as code (infrastructure, policy, configuration) is something an agent can work on. Today we’re extending that to everything after the merge with two launches: Pulumi Delivery, which lets you and your agents write delivery pipelines as code, and Pulumi Agents, customizable agents that watch your running systems and act when something goes wrong. Together, they take agentic code to production, and back again. I honestly can’t remember a launch since Pulumi 1.0 that our customers and team have been this excited about.
TypeSafe AI bet on Pulumi to scale Jev, their decision model, to over a trillion tokens per day using the engineering practices they already had.
What happens after the merge
Frontier labs have made huge advances in agentic coding, code review, and everything up…
As rapid adoption created fragmented access paths and inconsistent corporate governance, enterprise leaders formed a cross-functional taskforce to standardize large language model usage. The resulting infrastructure integrates directly with the existing data platform, Microsoft Entra ID, and Terraform to streamline oversight. Consequently, budget owners gain precise financial attribution and token tracking across all internal AI initiatives.
Azure Databricks has integrated its Genie One capabilities directly into Microsoft applications to streamline data access for business users. Grounded by business semantics and Unity Catalog controls, the rollout connects governed data directly into daily corporate workflows. This integration allows employees to query metrics and build assets without leaving their core productivity software.
Meta has transitioned its ads ranking ecosystem away from fragmented production models to resolve engineering bottlenecks and reliability issues tied to processing billions of daily requests. The company established Standard Model Templates to bundle architecture, feature logic, and serving configurations into reusable building blocks. This shift enables engineers to validate new methods on a small group of templates before rolling them out across the broader model fleet.