Introduction
Build browser and desktop workflows by demonstration, or compose APIs, code, and agents into general automations — then save reusable graphs.
What Is Ramain?
Ramain is a workflow-first automation platform with two complementary authoring paths. In Workflows, you describe what you want done while an agent drives a cloud browser or virtual desktop. A successful session compiles into a semantic workflow graph that you can save, edit, and relaunch.
In Agents, you chat with a builder that updates a mixed automation graph live. It can combine HTTP requests, code, branches, loops, LLM or agent nodes, files, approvals, and saved browser workflows.
Unlike brittle click-replay tools, Ramain prefers semantic agent blocks that reason about the interface, backed by defined inputs, workflow data, and reusable variables.
Workflow list overview. Current availability labels are Draft, Enabled, and Disabled; saving a valid graph makes it runnable.
Two Build Journeys
Workflows: browser or computer
- Create a workflow from Workflows home.
- Author in the workbench while the executor operates a live browser or desktop.
- Compile the latest trace into a graph on the Editor tab.
- Save the successful demonstration to compile an enabled workflow.
- Relaunch the saved workflow with new inputs.
Agents: general automations
- Open Agents and create an automation.
- Describe the process in builder chat while the canvas updates.
- Test in Sandbox, then publish the validated graph.
- Compose APIs, files, approvals, code, and saved workflows as child steps.
Every workflow and Agent automation uses draft or saved status. For browser and computer workflows, saving a valid graph makes it saved and enabled. Later edits autosave; there is no separate Publish step. Agents use their own Publish control.
Who This Is For
Key Concepts
What Makes Ramain Different
- Chat to graph: author by demonstration or conversational graph editing.
- Semantic blocks: agent tasks replace brittle coordinate replay.
- Dual runtime: automate web and desktop interfaces in one lifecycle.
- Composition: pin saved workflows inside larger Agent automations.
- SOP document: read and edit graph instructions with recorded screenshots.
Next Steps
Was this page helpful?

