AI Assistant · Automation
Automate recurring property operations with governed AI agents.
Define the objective, approved tools, trigger, records, approval points and expected result. AI Assistant can complete multi-step work while preserving human authority and session history.
Example agent
Review overdue PPM daily, group exceptions by cause, prepare owner follow-ups and send only after supervisor approval.
What are Automation Agents?
Automation Agents are the automation capability in the Platia 360 AI Features set. Each agent is defined before it runs: an objective, the tools it is allowed to use, what triggers it, which records it may touch, where it must stop for approval, and the result expected of it. That definition is the control, which is why an agent can complete multi-step recurring work - chasing overdue planned maintenance, reviewing vendor documents, preparing a daily portfolio brief - without human authority moving to the software. Every session is recorded and reviewable.
How it works
A governed workflow built around the operational responsibility.
Trigger
Start manually, on schedule or from a record event.
Reason
Determine what information or action is needed next.
Use tools
Read approved records or call connected systems.
Evaluate
Use the result to continue, stop or request input.
Approve
Pause sensitive actions until approval.
Record
Retain the session, tools, result and action.
Capabilities
Focused assistance without a disconnected chatbot.
Overdue work follow-up
Review exceptions, prepare messages and track response.
Vendor document review
Check required documents, expiry dates and evidence.
PPM exception management
Group missed work and prepare recovery actions.
Daily portfolio brief
Prepare a concise exception report from live records.
Enterprise controls
Source-linked, role-controlled and approval-aware.
AI works only with approved records and tools. Sensitive actions can remain draft, staged for review or approval-bound.
Permission-bound
Access follows the invoking user or configured agent role.
Auditable
Prompts, tool calls, results and actions can be retained.
Common questions
Questions buyers ask about Automation Agents
How does Automation Agents work?
Define the objective, approved tools, trigger, records, approval points and expected result. AI Assistant can complete multi-step work while preserving human authority and session history. In sequence: Trigger: Start manually, on schedule or from a record event. Reason: Determine what information or action is needed next. Use tools: Read approved records or call connected systems. Evaluate: Use the result to continue, stop or request input. Approve: Pause sensitive actions until approval. Record: Retain the session, tools, result and action.
What does Automation Agents actually do?
The capabilities on this page are: Overdue work follow-up; Vendor document review; PPM exception management; Daily portfolio brief. Each one operates on the same operational records the rest of the platform uses, so anything it produces is visible in the record's own history rather than in a separate AI log.
Does the AI act on its own?
No. It works inside the permissions of the user or the configured agent role, and actions that change a record or contact somebody are staged for approval rather than applied silently. Prompts, tool calls, results and the resulting actions can be retained, so any output can be traced back to who asked for it and what it was based on.
Where does the AI run, and does our data leave the environment?
That is a deployment decision made during scoping, not a fixed answer. Local and in-country model deployment is supported, so the operational records can stay inside the agreed hosting region. What is sent anywhere outside that environment is confirmed in writing before implementation begins.