Module — ACT
AI agents that act on your behalf. Always governed.
ACT is MarqX's AI action engine. It executes tasks — drafting responses, routing approvals, filling forms, triggering workflows — grounded in your policies, with human review gates on every high-stakes action.
80%+
of repetitive actions automated
100%
actions traceable to source policy
< 2 min
avg agent response time
What ACT does
AI action execution with governance built in
Policy-grounded execution
Every ACT agent action is grounded in documents from KNOW. No hallucinations — every output cites the policy it followed.
Configurable human gates
Set thresholds for which actions require human approval. Low-risk actions auto-execute; high-stakes ones queue for review.
Multi-channel deployment
Deploy ACT agents in email, chat, WhatsApp, API, or internal ops tools. One agent, every surface.
Workflow triggers
ACT agents can trigger RUN workflows — starting approval chains, creating tasks, or updating records based on action outcomes.
Full action log
Every ACT output — draft, approval, execution, override — logged with agent version, policy cited, and reviewer identity.
Parallel agent execution
Multiple ACT agents run concurrently across different workflows. Scale without adding headcount.
Deploy AI agents your compliance team will approve
See ACT handle your highest-volume action flows — support responses, approvals, or data entry — with citations and human oversight.