tableline. Built for hospitality

YOUR NEXT GREAT HOST IS ALWAYS HERE

Busy tables.
Never a missed welcome.

Let your team take care of the room. Your AI host takes care of the call.

Natural conversation.
A reservation, handled.

CK
YOU'RE CALLING

The Copper Kettle

Checking setup
+ A LITTLE MORE HUMAN
VOICE, WITH PURPOSE +
Meet your AI host AI

Good conversations start here.

Ask for a table, explore the menu, or plan your visit.

Microphone on only during a call
CALL DURATION00:00
REPLY WAIT — ms
YOUR TURNS00

LESS HOLD MUSIC.
MORE HOSPITALITY.

One conversation.
Three things off your plate.

Fill tables, effortlessly.

Checks availability, confirms the details, and returns a booking reference.

Reservations, handled

Know the menu by heart.

Answers from the menu, with prices, dietary preferences, and dish details.

Answers with context

Keep it conversational.

Speak naturally. Change your mind. Interrupt when you need to.

Your pace, your language

A SMALL TEAM'S EXTRA PAIR OF HANDS

Hospitality shouldn't
come with a hold queue.

When a restaurant is busy, the phone competes with the people in the room. Tableline gives callers a voice host that can answer questions and take reservations.

  1. 01
    Make the ask.

    Try a table for four tomorrow at 7. The host checks the actual reservation tool.

  2. 02
    Change your mind.

    Interrupt the host or ask about vegetarian dishes. Watch the transcript and action results.

  3. 03
    Leave with an outcome.

    Agree to the booking. Your confirmation appears beside the conversation.

Choose your demo mode

Live conversation connects your microphone to AssemblyAI through this server. The provider processes audio and may retain session recordings and transcripts.

Interactive demo is a scripted walkthrough with simulated actions. It never opens your microphone, contacts the voice API, or creates a reservation.

This prototype runs browser calls; it is not connected to a telephone number. This build uses an in-memory reservation store. Live bookings last only for the server process and are not shared between instances.

A MEASUREMENT, NOT A PROMISE

What “reply wait” means.

Measured in your browser from receiving the speech-stopped event to scheduling the first reply audio for playback. It includes the playback cushion, but excludes speech detection before that event and device output latency.

The number is the latest measured reply, not a full end-to-end benchmark. Interactive demo timing is never reported as live latency.