Budget
Crosses execution vs BOQ at end of day. Alerts before overrun.
The agents
Agents that work on your project's real context
It's not a chatbot. These are specialized agents that consult your project data, propose actions and act under the policies and approvals you define — every step is audited.
The agents
A governed workforce, not blind autonomy.
Each agent specializes in one part of the flow. They consult your data, propose actions, and act under policies and approvals you define.
Evidence
Classifies photos by chapter, validates quality, tags risks, and geolocates.
Suppliers
Prepares orders, validates invoices, and reconciles payables.
Schedule
Replans when an activity slips more than 48 hours.
Commercial · CRM
Tracks opportunities, reminders, and WhatsApp replies with templates.
Reports
Generates the weekly site report ready to send to the client.
Governed, not blind autonomy
AI as an applied capability, with guardrails. You decide how much autonomy to give each agent.
Human approval on sensitive steps
You define what each agent can do alone and what needs your sign-off before it runs.
Least privilege
Each agent only accesses the sources and actions it needs for its task, nothing more.
Full audit trail
Every run is logged: what it did, which sources it used, its cost and whether it went through approval.
Isolated per company
Agents only see your company's context; data never crosses with anyone else's.
Frequently asked questions
Do agents act on their own?
No, not blindly. Each agent proposes and, per the rules you set, automatically runs routine work and leaves sensitive steps for your approval. Everything is audited.
What data do they work with?
Yes. Your project context: BoQ, schedule, evidence, observations and closed projects. They adopt your way of working, not a generic one, and only within your company.
Can I audit what each agent does?
Yes. Observability logs every run: what it did, which sources it used, its cost and whether it required human approval.