Build governed AI workflows. Visually. Or in four lines of Python.
Visual workflow builder
A workflow with a human approval gate, a connector step, and a policy check — built in minutes, not weeks of bespoke integration.
Two paths, one result
However you work — canvas or code — you land on the same governed execution: provenance, policy, and human gates when stakes demand it.
| Visual | Code |
|---|---|
| Drag a step | import human |
| Add a gate | @human.workflow |
| Deploy | async def run(ctx): … |
| Governed. Running. | Same ledger-backed run. |
Deeper narrative: Four lines of Python — protocol enforced.
Workflows are not automation
Automation that removes humans is a different category. HUMΛN workflows keep people in the loop at the moments that matter — from personal usefulness to enterprise-wide standards — without shrinking work to a single segment.
Universal lifecycle
- Useful immediately — individual value
- Repeatable — habit and routine
- Customizable — builders adapt it
- Shared — the team relies on it
- Governed — formalize when stakes rise
- Standardized — enterprise-wide
Launch a workflow · Build a repeatable flow · Share it · Add governance when accountability matters.
Every step produces a governed artifact. Inspect any of them in the Console — executions link artifacts back to policy, identity, and provenance.
View executions and artifacts →Example starting points
Governed outputs
Every step can emit a structured artifact — signed, inspectable, and contestable — with provenance behind it. That is the bridge from execution to governance.
Live canvas (screenshot)
