How a governed AI agent actually works
The agent never touches your systems directly. Everything goes through a gateway you control. Play the three scenarios from the demo and watch where each one is allowed, blocked, or held for you.
Read project data
Try to reassign people
Draft the recommendation
Clear
Gateway: the agent only sees this
Decision owner sets the boundaries, signs the decision
AI agent reads, reasons, drafts asks the gateway for everything
Identity and permissions who the agent is, what it may touch
Budget and run limits € per run, max steps, time
Tool allow-list what it may call; the rest does not exist
Approval hold actions with impact wait for a person
Audit log every call, outside the agent's reach
Data sources project, history, capacity, PTRS
Language model writes the narrative, does not classify
Rules engine classifies risk, cites the rule
Tools with impact change gate, move people, send mail
Decision record what was allowed, what happened, who signed, what it cost
request
escalate / approve
Click a box
Each box is one thing the agent depends on or one control the agent cannot bypass. Click one to see what it does in plain language, or play a scenario to watch a call travel through the map.
Ready.
allowed blocked waiting for you