A shared picture.
A better release.

Built for the whole delivery conversation

• DevOps and automation engineers: connect infrastructure, build, deployment, verification and recovery in a single model.
• Developers: understand dependencies, database changes and the path from a source change to production.
• Architects: review environments, runtime dependencies, delivery decisions and target designs with a team.
• Product owners: see where manual work and uncertainty affect delivery, then agree on an improvement backlog.
• Consultants and clients: use a shared diagram to surface assumptions, discuss alternatives and capture what was agreed.

Run a useful design session

1. Choose a starting point. Open a template, explore the example workshop, start a blank blueprint or load an existing file.
2. Map the current design. Add your environments, components and activities. Describe what is manual, assisted, automated or unknown. Record owners, assumptions and evidence where you have them.
3. Connect the work. Show execution prerequisites, promotion between environments and runtime dependencies. Include verification and recovery so the diagram explains more than the happy path.
4. Design a target scenario. Clone the current scenario and change the proposed process. Keep today's reality available while you compare what improves.
5. Agree on next steps. Review findings and open questions. Capture improvement items, ownership and acceptance criteria for the team to discuss.
6. Save and share deliberately. Download a blueprint file to continue later or share it with colleagues. Export diagrams and process documentation for your workshop output.

Example: an automated app, a manual database

Your app deploys through a pipeline, but an operator runs database scripts and checks production by hand. Model those activities as they work today. Capture the database dependency, who runs the scripts and how success is verified.

In the target scenario, propose a controlled migration step, a verification gate and a recovery path. Discuss ordering and compatibility before connecting parallel work. The result is a design the client, product owner and engineers can review together.

Work with your AI assistant in both directions

Choose Export → Work with AI… in the editor to prepare a prompt and blueprint for your assistant. Ask it to map the design into implementation tasks, tests and open questions. You choose when and where to share that handoff.

You can also give an assistant workshop notes or feedback and the public schema, then ask for an .automationmodeler.json blueprint. Open the generated file in Automation Modeler, review the model and refine it with your team.

The app supports this exchange through files. It does not run an AI model, deploy infrastructure or execute the modeled process for you.

Your blueprint stays under your control

Blueprints are stored in your browser and in files you choose on your computer. Save a file before clearing browser data or moving to another browser or device. An AI assistant receives your blueprint only when you share it with that assistant.

The optional usage counter stores daily totals for app opens and blueprint creation. It does not store visitor or blueprint details. The app's Privacy panel explains the counter and its controls.