The questions we get most often.

Hivematic is a platform to build and run autonomous AI agents that operate on the real processes of your business. It's not a chatbot: agents receive events (tickets, emails, webhooks), decide on their own and act on your systems — with a Guardian supervising every critical action.

Three things: (1) agents act, they don't just chat — they execute real actions on your systems; (2) everything is in the platform — RAG, ontologies, audit trail, escalation, monitoring — you don't assemble 12 SaaS; (3) it's built for production: Guardian AI, simulation mode, full audit. Security-first.

An agent follows a clear pipeline: it receives an event (email, ticket, webhook, scheduled trigger), decides which task to run via AI routing, executes the configured actions (reply, classify, update a CRM…) and closes. Everything is tracked in the audit trail.

Yes. Each agent has a configurable persona: tone of voice, scope of expertise, system prompt, knowledge sources. You can have a polite, formal "Customer Care" agent and a more direct, technical "Engineering" agent — on the same platform.

Workflows orchestrate end-to-end multi-step processes. You combine AI steps, conditions, human escalations and actions on third-party systems in an executable graph. Perfect for processes that need more than one decision: multi-level approvals, lead qualification with follow-up, order reconciliation.

Native features ready out of the box: omnichannel inbox, contact management, knowledge base with RAG, ontology engine, funnel & pipeline, scheduler. You don't have to integrate an external CRM to get started — everything is in the platform from the Free plan.

Yes. Simulation mode runs the agent as if in production, but intercepts all output actions: you see what it would have done without actually touching email, CRM or third-party systems. It's the only reliable way to validate an agent before launch.