Four shapes Dock isn’t.
They’re the four default ways teams put AI to work today, and all four break at the same place. Here’s the shape we built instead.
Not a chatbot. A shared workspace.
Not AI bolted onto a doc. Agents as first-class identities.
Not a stateless agent. The workspace as the substrate.
Not a single-player tool. Plural by default.
What an agentic workflow looks like, plainly.
An agentic workflow is a sequence of actions an AI agent takes on its own to complete a task. Read a row, decide, call a tool, update a row, hand off, repeat. The agent is the worker. The workspace is the place the work lands.
Most agentic workflows fail not because the agent reasons poorly, but because there is nowhere for the work to live in a way the next agent or the next human can see. The workflow exists. The substrate doesn’t.
On Dock, the substrate is the workspace itself. Typed tables for structured work. Docs for prose. Comments for review. Real-time presence. Every action signed by the principal that ran it. The agentic workflow is just a sequence of writes against a shared surface that humans and other agents can read in real time. It’s the shape we think the next era of the cloud takes: Cloud 2.0.
See the patterns in practice in “How humans and AI agents actually work together” and the architecture in “Agentic AI architecture: the five layers nobody draws together”.
One workspace. Humans editing in the UI. Agents calling MCP. Everyone on the same row.
An agent rewrites a row. The version history names the agent. Another agent picks up the next item in the queue. A human reviews both edits and approves. Nothing leaves the workspace. No screenshots flying through chat. No “who wrote this” guesses.
That’s the shape. If you want the full thesis, read the manifesto. If you want the engineering, read the blog.