Case study, featured from the Inge de Bruijn Zwembad platform

146 customer questions, answered automatically.

At Inge de Bruijn Zwembad the same questions about lessons, rates and times came in all day, while the answer was already on the website. I built an AI receptionist that handles that volume, day and night. It's part of the larger platform I built for the pool.

146customer questions handled, excluding test data
3,454conversations on the live counter, and it keeps going
Secondsfrom question to answer, not days
24/7available, also outside opening hours

Phones ringing off the hook, while the answer was already online.

What time does the lesson start, what does a ten-visit pass cost, is the pool open during the holidays. The same questions every day, by phone and email, while the answer was already on the website. The team lost hours a week to it, and outside opening hours everything just sat there.

  • The same questions over and over about lessons, rates and times
  • Phones ringing off the hook at peak times
  • The answer was already on the site, but nobody found it
  • No answers outside opening hours

What I built.

A receptionist that takes the volume, with the team in the driver's seat.

01

Its own knowledge base

Rates, lesson times, rules and common questions in one place. That's the only source the receptionist answers from, so there's no guesswork.

02

AI receptionist, 24/7

Answers customer questions within seconds, at 23:00 and on Sundays too. The same approach runs in several languages at Referee Abroad.

03

A person as soon as it gets tricky

When in doubt, or with complaints or sensitive questions, the receptionist passes the conversation to a staff member. AI takes the volume, people handle the custom work.

146 questions no person had to answer anymore.

146 customer questions handled, counted without test data. The platform's live counter now stands at 3,454 conversations and keeps climbing while you read this. The team stays in control and picks up the conversations that really matter.

under the hood A language model on its own knowledge base, handover to a staff member, and a counter that comes live from the platform. Fully documented, and the client owns the code and documentation.

Want to be rid of the same questions too?

An AI receptionist can also be built on its own, without the whole platform around it, as a quick win from €750. In the free process review we check whether it pays off for you.

Walk me through one process