AI Agents · ~1 credits / run
Conversation / AI Reply Agent
The Conversation / AI Reply Agent drafts on-brand replies to DMs and comments using your Knowledge Base. It classifies each message's intent — lead, support, or spam — and never invents facts: when it isn't sure, it escalates to a human instead of guessing.
What it does
- Drafts on-brand replies to DMs and comments from your Knowledge Base.
- Classifies intent: lead, support, spam, or other.
- Flags genuine leads so you can follow up fast.
- Escalates to a human whenever it's unsure — never fabricates answers.
- Keeps replies consistent with your brand voice.
Example output
markpilot.com — Conversation / AI Reply Agent
Drafted an on-brand reply using the Knowledge Base. Classified intent as a lead; suggested capturing email + offering a sample.
- → Classified intent + sentiment.
- → Retrieved context from KB.
- → Drafted reply in brand voice.
Brand-safety check passed · 91% confidence
Who it's for
- A brand drowning in repetitive Instagram DMs and comments.
- A solo founder who wants leads in the inbox surfaced instantly.
- Any business that wants fast, on-brand replies without canned templates.
Frequently asked questions
- Does it reply automatically?
- It drafts the reply and classifies intent; you stay in control of sending. Anything it's unsure about is escalated to you.
- How does it know what to say?
- It uses your Knowledge Base, so answers reflect your real policies, offers, and facts — not generic guesses.
- Can it spot leads?
- Yes — it classifies each message and flags genuine leads so you can prioritise them.