AI & Automation
Support doesn't scale by hiring alone. Quackback pairs an AI teammate named Quinn with a trigger-driven workflows engine, so routine replies and routing happen without a human in the loop every time.
Everything lives in Admin → Settings, under AI & Automation: Agent, Copilot, Skills, Connectors, and Workflows.
Quinn needs an AI model configured. The Agent replies in Messenger once the support inbox is on. Workflows appear when the support inbox is on.
Quinn: two peer agents
Quinn isn't one bot with two names. It's two separate configurations that share the same identity:
- Agent replies to customers directly in the Messenger widget. It works the moment you deploy it.
- Copilot sits beside your team in the inbox: it answers questions, drafts replies, and never speaks to a customer directly.
Both agents draw on the same knowledge sources, though what each can see differs. Both can run actions, including tools from MCP servers you add as connectors.
Workflows
Workflows automate the parts of support that don't need judgment: routing a new conversation to the right team, applying an SLA by priority, closing a ticket after a week of silence. Workflows are built from triggers, conditions, and actions, and they work on both conversations and tickets.
Workflows ship with a template gallery so you don't have to start from a blank canvas, and a library of reusable conversational blocks for building interactive Messenger flows.
Monitor
Once Quinn is live, the Quinn section of Admin → Analytics shows resolution rates, handoffs, action outcomes, and how much your team uses Copilot. See Monitor Quinn.
Next steps
- Configure the Quinn Agent: deploy Quinn to Messenger and set its knowledge and guidance
- Build a workflow: automate routing, SLAs, and housekeeping
- Give Quinn actions: let Quinn close conversations, create tickets, and use your own tools