Title here
Summary here
Meet Monty — Your AI Messaging Assistant
Your inbox, pre-sorted
More than half of all customer messages are about planning — accepting a proposed date, asking to move, cancelling attendance. Another quarter need no action at all (“Thanks!”, a 👍). Until now you had to open every conversation to find that out. Monty, the new AI assistant in the messaging panel, does that first pass for you.
Colored tags on every incoming message
Every incoming customer message is automatically categorised with a small colored letter, in the conversation list and on the message bubble:
| Tag | Meaning |
|---|---|
| P (blue) | Planning — date accept/decline, reschedule, availability, cancellation |
| Q (teal) | Question — practical info, booking changes, vouchers, payment |
| W (teal) | Weather — “will the flight go ahead?” |
| C (red) | Complaint or sensitive situation — always needs a human touch |
| N (grey) | No action needed — thanks, acknowledgements, plain emoji |
Scan the list by color: grey rows can be skipped, red rows get your attention first, blue rows are scheduling work.
Reply drafts you review — Monty never sends anything
For planning messages, Monty drafts a reply in the customer’s language, grounded in the actual booking, flight and location data. The draft sits pre-filled in the composer under a clearly marked Monty strip — read it, adjust it if you like, and press the normal send button. Sending is always your action; dismissing the strip discards the draft.
When a message implies a booking change (“Tuesday works for us!”), Monty also shows a compact action card: Confirm booking on flight Tue 21 Jul — evening (3 p.) with an explicit Accept button. You can switch to a different flight before accepting, and the change runs under your own account with all the usual capacity and status checks. An orange dot on the conversation row tells you a proposal is waiting.
Monty learns how you answer
In Settings → Monty you can write instructions per category — your tone, your standard meeting-point info, your policies. Even better: press Learn from history and Monty reads your past conversations (you pick 1, 3, 6 or 12 months, with the message count and estimated cost shown up front) and distills how you typically answer into an editable instruction draft per category. Adopt it, tweak it, or discard it — from then on, proposals sound like you.
Your key, your costs, full transparency
Monty runs on your own Anthropic API key — entering it in Settings → Monty is the opt-in, and all usage is billed to your own Anthropic account. The settings section shows per-month counts and an estimated cost in € (typically well under €0.05 per handled conversation). If your credits run out, Monty simply pauses — the messaging panel keeps working exactly as before, and Monty resumes automatically once credits are back.
See the Monty settings guide for the full setup walkthrough.
Happy flights!