A Telegram-native appointment booking system. Customers chat the bot, share their phone with one tap, get a queue number. Behind the bot, the business runs a dashboard for staff schedules, today’s bookings, and the live “now serving” counter.
See it live → chat the Pinlon booking bot on Telegram. Ask “who’s working today?”, “book me Saturday”, or “what’s the queue right now?”.
Powered by HI — every business is its own tenant with isolated data and its own bot.
What customers get
- Chat the bot in Burmese, English, Thai, or any other language — replies match the customer’s language
- See the current queue any time: “Now serving #12, 4 people waiting”
- Verified phone via Telegram’s Share Contact button — single tap, no SMS code to type
- Queue number assigned in booking order (FIFO) — simpler than picking time slots
- My bookings lookup by phone or Telegram ID
What admins get
- Live appointments dashboard with status workflow: Booked → Checked In → In Progress → Complete
- One-tap “Update serving number” — auto-advances on Complete
- Cancel with reason vs hard delete for typos, kept distinct for reporting
- Bot conversation log with thumbs up/down feedback, so prompt issues can be fixed iteratively
- Broadcast to all past bot users with one click — Markdown messages with optional image
Two queue modes per business
- Serial (FIFO) — strict one-at-a-time. Best for clinics, single-chair barbers. The system rejects starting a second customer until the first is complete.
- Parallel — multiple in service simultaneously. Best for salons with multiple stylists, workshops with multiple bays. No concurrency check; admin starts whoever they want.
One radio button in business settings switches between them.
First customer in production
A beauty salon in Yangon (“Pinlon”) is using the system live: two stylists, daily walk-in flow, Telegram bot taking bookings around the clock. We watch the bot conversation log to spot prompt issues and tighten rules.
Want it for your business?
Reach out — we’ll set you up with your own Telegram bot and dashboard. Onboarding is about 10 minutes (we paste your bot token, add your staff, and you’re live).
Tech
Next.js (admin), Postgres (Neon), n8n (Telegram workflow with AI agent), MCP server (7 booking tools), Telegram Bot API. Self-hosted on a single VPS, deployed via PM2 + Docker.
Updates
What's shipping, what's fixed, what's next.
-
Booking is live with first customer
Telegram-native appointment booking is live. First production customer is a beauty salon in Yangon — bookings flow through the bot, phone verified by Telegram's Share Contact, queue auto-advances as customers complete.
The first version of Booking went live this week.
What shipped
- Multi-tenant platform — each business gets its own admin dashboard, Telegram bot, and isolated data
- Telegram bot via n8n — gpt-4.1-mini agent with 7 MCP tools, multilingual replies, conversation memory keyed on Telegram user ID
- Free phone verification — Telegram Share Contact button replaces SMS OTP entirely. One tap, cryptographically verified by Telegram.
- FIFO + parallel queue modes — same database, one radio toggle per business
- Bot conversation logs — admin sees every reply, tags good/bad with notes, used to iterate the system prompt
- Broadcasts — push a Markdown message (with optional image) to every past customer in one click
- Cancel-with-reason vs hard delete — distinct destructive actions, kept clean for reporting
Production customer
A beauty salon in Yangon (Pinlon) is the first live business. Two stylists, parallel queue mode, Telegram bot taking real bookings.
What’s next
- Reminders the night before the appointment
- No-show tracking with optional deposit gate
- Calendar export to add bookings to phone calendars
- Public web booking page (Calendly-style URL) sharing the same data