Everything Herrald does

Herrald gives you one calm inbox for every comment and DM across your brands. It reads everything, drafts replies that sound like you, and makes sure the hard conversations reach a person.

Works with

Facebook
Instagram
YouTube· Coming soon
X· Coming soon
Your own app

What should come next?

One inbox for every conversation

Every network in one place: Facebook and Instagram work today, and YouTube and X are on the way. Your own app can connect through our signed integration kit for Laravel and Express.
Many brands, one workspace: Each brand keeps its own accounts, voice, knowledge and rules.
Comments by post: Threads gather under the post they’re on, with its picture, caption and a link to the original. Prefer one long list? That’s a click away.
Most important first: Possible crises come first, then failed replies, closing DM windows, questions and comments on your ads. Thank-yous wait at the bottom, and every row says why it’s there.
Search as you type: Find any comment or person by a word or a name. Results show up while you’re still typing.
Views with live counts: Built-in views show what needs you, what’s yours and what nobody has picked up yet. You can save your own too.
One clean filter bar: Filter by brand, network, status, comments or DMs, assignee, post type (organic, boosted or ad) and reason.
Live without the jumping: New conversations wait behind a “3 new” banner instead of shoving the list around under your cursor.
Your own focus: Hide brands or networks from your inbox without hiding them from the team.

AI triage that knows when to step back

Every message understood: Each message lands in one of 10 categories, from thank-yous and simple questions to order problems, disagreement, spam and possible crises. You see its mood, risk and confidence, explained in plain words.
Drafts in each brand’s voice: Drafts use the brand’s facts and the replies you keep as examples, so they read like your best answers.
Guardrails on every draft: Every draft passes 9 checks before anyone sees it: allowed links only, no order or money promises, no medical or legal advice, length, same language, no third-party mentions, no echoed instructions, no placeholders and crisis words.
A second opinion: A separate safety check takes another look before anything goes out on its own.
Crisis-aware: Possible crises go straight to your owners with support resources ready, and they’re never answered automatically.
Stays out of arguments: It won’t draft replies to comments aimed at another commenter.
Built against manipulation: Message text is only ever read, never obeyed, so a comment can’t give it orders.
On demand: Tap “Ask Herrald” for a fresh reading, or “Draft a reply” when you’d like a draft.

You decide what happens

Rules per brand and category: For each kind of message, you choose: send automatically, draft for review, review without a draft, escalate, ignore or hide.
Evidence before automatic replies: Automatic replies stay off until you turn them on. First, Herrald replays your recent messages and shows you what it would have sent.
Safety controls: There are off switches for the whole workspace and for each account, plus quiet hours, hourly and daily caps, and a safety margin before reply windows close.
Workspace defaults, brand overrides: Set your defaults once, in your own time zone, and let any brand do its own thing where it needs to.
Versioned settings: Every change is saved with a note, so you can read the history and go back in one click.
Quiet the noise: Mute a heated post and all its threads at once, and let old posts mute themselves.

Built for a team

Roles and scopes: Invite people by email as Owner, Admin, Moderator, Reviewer or Viewer, and choose which brands and actions each one gets. Removing someone takes effect right away.
Queue mode: Work through a view one conversation at a time, with keyboard shortcuts.
Bulk actions with Undo: Assign, take, escalate, close or reopen one conversation or a whole pile. You get Undo instead of “are you sure?”
Two people, one thread: You can see who’s reading or writing, and Herrald speaks up before two people answer the same person.
Never miss urgent work: You’ll get an email when something’s assigned to you, a reply fails or a conversation is escalated. Everyone picks their own alerts, and a live count sits in the sidebar and the browser tab.
Phone-ready: A thumb bar keeps the recommended action and the reply box within reach.
Permissions you can see: If a button is greyed out, it tells you why in plain words.

Handle every conversation well

A recommendation card: Herrald suggests what to do next and tells you why.
Replies in the thread: Replies land right under the comment you’re answering, and you can like or hide comments as the Page.
Reply windows: Herrald knows each network’s window for DMs and tells you when one is closing.
Failed replies come back: Herrald tries again, and if a reply still fails, it comes back to you with a “Send again” button.
Edits and removals tracked: If someone edits a comment, you’ll see each version, and removed comments never get a reply.
Post context: Every thread shows the post it’s on and whether that post was organic, boosted or an ad.

A home for every brand

One page per brand: Each brand’s profile, accounts, voice, knowledge, rules, quality and costs live together on one page.
A readiness checklist: It tells you what to set up next.

Costs you can act on

A usage meter: You’ll get a heads-up at 80%. At 100%, drafting pauses until you approve more, but reading and sorting carry on.
Spend where it happens: See what each post, category and step costs, right next to the setting that changes it.
Bring your own key: On Bring your own key, every AI call runs on your own Anthropic key, and you set the monthly cap.

Trust, security and privacy

Workspaces sealed off: Each workspace is isolated in the database itself, and the app never runs with database admin rights.
Encryption: Account tokens and API keys are encrypted with each workspace’s own keys, and those keys rotate. Sessions are encrypted too.
Sign-in protection: Owners and admins sign in with a passkey or an authenticator app. Reviewers and viewers can use an emailed sign-in link. Single sign-on from your own platform works too.
An audit log nothing can change: Every action is written down as a plain sentence, and you can export the lot as a CSV.
Security alerts: You’ll get an email the moment something changes how your workspace signs in or replies.
Your data, your call: Export everything, delete it with a grace period, and choose how long old conversations are kept.
Hardened defaults: A strict content security policy is always on, secrets are masked in logs, and nothing sensitive is left in links or browser history.

Runs itself

Workspace health: Your accounts and background jobs are checked all the time. If something needs fixing, it tells you what and where.

Want to be one of the first?

Herrald is opening to a few teams at a time. Leave your email and we’ll write as soon as there’s a spot for you.