Skip to content

Inbox agent · Local Businesses

Instagram & TikTok DM replies for local businesses

The Inbox agent drafts DM and comment replies in your voice — for opening-hours questions, “do you have…?” stock questions and booking and reservation requests — and queues them for one tap. Automatic sending is off by default.

Small dark modules connected to a central dotted sphere, one of them lit blue

The problem

Local Businesses lose enquiries in the DMs — or answer them wrong.

The same questions arrive every day: opening-hours questions, “do you have…?” stock questions. Answer slowly and the person moves on; answer with a canned auto-reply and it sounds like nobody's there.

Keyword-triggered DM automation makes it worse when it misfires. For local businesses, a stray reply about prices is a real problem — which is why the Inbox agent drafts and you send.

IllustrativeDrafted caption · fictional example, not a customer

The one thing regulars order that nobody else knows about

It's not on the menu board, and it never will be. Ask for it by name next time you're in — we'll know you saw this.

Signal · Behind-the-counter clips got more saves than product photos this month.

How it works

What the Inbox & approvals agent does, step by step.

Local Businesses on Instagram and TikTok. Every step ends in a draft, not an action.

  1. 01

    Read the thread

    The agent reads each new DM and comment with the context of the conversation so far.

  2. 02

    Draft from your brain

    Replies use your offers, FAQs and friendly and local, like the person behind the counter — and flag anything touching prices.

  3. 03

    Queue for one tap

    Drafts wait in your inbox. Approve, edit or skip — on your phone between everything else.

  4. 04

    Learn reply skills

    Replies you approve become reusable reply skills, so routine answers get better over time.

What you get

Drafts you can use, with the reasons attached.

Part of the Inbox & approvals side of Oryxia, reading from the same brand memory as the other six agents.

  • Drafted replies for opening-hours questions, “do you have…?” stock questions, booking and reservation requests
  • Comment replies queued alongside DMs in one inbox
  • Flags on anything sensitive, uncertain or off-brand
  • A library of reply skills learned from what you approved
Draft · awaiting your approval

Every reply waits for you. Auto-send is off by default.

Even if you switch on routine replies later, the agent can only send answers you approved in advance — anything uncertain still waits for you.

  1. 1One-tap approve, edit or skip
  2. 2Automatic sending off by default
  3. 3Uncertain or sensitive threads always wait for a person

A good fit if

  • You get repeat questions — opening-hours questions, “do you have…?” stock questions
  • You've been burned by, or are wary of, DM auto-responders
  • You want to reply fast without handing over judgment

Not a fit if

  • You want a bot to run high-volume giveaway or comment-to-DM funnels with no one watching — the Inbox agent is built the other way round.
  • You want someone on site to film for you — Oryxia works with the footage you shoot on your phone.

FAQ

Fair questions.

01Can it reply automatically?

Automatic sending is off by default. You can switch on routine replies, but even then the agent can only send answers you approved in advance. Anything uncertain waits for you.

02What happens with sensitive messages?

They're flagged and never auto-sent. For local businesses, that includes anything touching prices, stock and offers that aren't current. You decide how — and whether — to answer.

03Will it understand how a business like mine talks to customers nearby?

It starts from a 30-minute brand interview and your own past posts, aiming for friendly and local, like the person behind the counter. Your do-not-say list — for local businesses, usually prices, stock and offers that aren't current — is enforced on every draft, and you approve each one.

See what your agents would draft first.

The free Social Audit maps your niche, competitors and hook gaps, and shows the agents working on a demo workspace.