I built a Cloud Engineer Agent so app teams stop waiting on us. These are Enterprise production workloads.
New workloads used to mean a custom AWS build and a ticket pile. IT, Security, Networking, and FinOps showed up at the end. Standards drifted. We were the bottleneck.
Now they state the workload. Agents write production Terraform against a platform template. A human merges.
Hours. Not a project.
Cloud scaled. Adoption scaled. The ticket queue didn't.
We built a Cloud Engineer Agent so app teams stop waiting on us. They describe the workload. Agents write Terraform against the platform standard. A human merges.
Hours, not a project. If AI doesn't change how the company ships infrastructure, it's a demo.
I built a Cloud Engineer Agent so app teams stop waiting on us. These are Enterprise production workloads.
New workloads used to mean a custom AWS build and a ticket pile. IT, Security, Networking, and FinOps showed up at the end. Standards drifted. We were the bottleneck.
Now they state the workload. Agents write production Terraform against a platform template. A human merges.
Hours. Not a project.
@elonmusk can you allow me to use cursor on mobile where I can go hands free? I want to prompt with my voice and then I want to hear the results of periodic updates or when the task is finished through headset or Bluetooth
This would allow me to work as if I was Tony Stark working with Jarvis
Going to Disneyland tomorrow. Using @bot to route my day for my daughter AND I’m using it to automatically sign up for my fast passes based on our group preferences.
Told a cast member my plan and they were like idk how I feel about that… told them to download grok and connect their grocery app, give it a budget and some ideas of what they like to eat and POOF it will be at your front door
@Joshua_WD Thanks Josh! When deploying to prod and spinning up cloud resources that cost $$ - you need to have a human review and be the gate for deployments
@vinvan Interesting take. I do like this approach because ultimately I’m curious about features, flows and architecture not necessarily the code.
I’ve built a GitHub action into my apps to self document and push these diagrams into the read me or docs through the repo
How it runs: request → plan → generate → MCP read → PR → plan, lint, policy, cost. Fail: iterate. Pass: a human ships. Lessons go back into the template, so the next app starts stricter.
The business result: cloud scaled, adoption scaled, the ticket queue didn't.
If AI doesn't change how the company ships infrastructure, it isn't transforming anything. It's a demo.
What it is: five specialists on one change. Plan, write Terraform, validate, document, review.
What it solves: every team reinventing networking, encryption, IAM, and CI. They fork a platform template, so the secure foundation is already there. Control-team rules ride in the context pack, not in a late review.
MCP reads live AWS and current Terraform schema. Git is the only write. The PR is the product.