AI chatbot for customer service

Answer customers at 2am, from what your systems actually say.

A chatbot connected to your live records rather than a page of FAQs. It answers from the data, hands over to a person where you decide, and says it does not know instead of inventing.

The situation you are probably in

The same questions arrive all day and half the night. Where is my order, what does it cost, what time does it start, what do I still owe. One or two people answer them by hand, the answer changes with whoever is on shift, and anything asked after six waits until morning. The customer who waits asks someone else.

What we do about it

We wire the chatbot to the systems that hold the answers, so it reads live records instead of a document someone wrote six months ago. Then we grade it the way a customer will use it: a test set of real questions with known answers, including the ones where the only correct reply is to refuse or fetch a person. It ships with its limits written down, and every conversation lands in one log your team can read.

Typical engagement. Three to six weeks from scope to a chatbot answering customers in production, then a monthly retainer to run it, widen what it can answer and re-run the test set.

Where this already runs

Built for a private multi-day event running across two countries: a concierge on the web and inside the guests' own messaging app, answering from the live plan rather than a copy of it. When it updates the plan it uses the same path a person does, so there is one audit log. It scored 40 out of 40 on a test set of real guest questions, including seven where the correct answer was to refuse. The same shape answers market questions inside a messaging app for our own market intelligence product.

Most builds of this shape start with the two-week assessment: we map the process as it runs today, cost it, and write the scope. The fee is credited in full against the build.

Start with the two-week assessment

What you receive.

Concrete things, not outcomes on a slide. This list is the scope of the engagement.

  1. 1

    A chatbot on your site, and in the messaging app your customers already use, answering from your live data

  2. 2

    A written boundary: what it answers, what it hands to a person, and what it must never say

  3. 3

    A graded test set of real customer questions, refusals included, re-run on every change

  4. 4

    Handover to a named person at the points you choose, with the conversation attached

  5. 5

    One log of every conversation and every change the chatbot made, readable without an engineer

  6. 6

    A handover session so your team can change its answers without us

A good fit when

  • The same questions arrive twenty or more times a week
  • The answers already live in a system, even if that system is a spreadsheet
  • Someone inside the business can own the answers after launch

Price it before you commission it

Put the hours the task takes today, the salary of the person doing it and any quote you hold through the payback calculator. It shows the yearly saving, the payback period and the arithmetic behind both. Free, and nothing is stored.

Run the payback numbers

Things you can check.

Figures from our own systems that this service draws on. Each one names its source.

40/40
guest questions answered correctly, including the ones set to trip it up
120+
tables of business data you could export tomorrow
25+
business systems already connected: CRM, email, ads, payments

Every figure names its source. How we count. The investing system runs on a broker paper account, not real money.

Make the next AI project one the business can measure.

Thirty minutes with the founder, no slides. You will know what the right solution looks like, what it would take to build, what it should return, and which part to start with.

Replies come from a named person in Dubai within one working day.