AI assistants and chatbots

AI that uses your company data and carries out specific tasks

I build assistants that answer questions, check data and support work in the systems you already use.

Let's discuss your AI assistant
Illustration of an AI assistant connected to company data

Sounds familiar?

Your team should not answer the same questions over and over

Customers and staff ask about the offer, orders, dates and documents. The answers sit in files, inboxes or one person's head. Without them, everyone has to wait.

The same questions come back every day

What does it cost, what is the status, when will it be ready, where is the document. Someone on the team replies by hand, even though the answer has already been given before.

Company knowledge is scattered across documents and inboxes

The offer, FAQ, procedures and exceptions sit in files or emails. To reply, someone first has to find the right person or the right document.

A generic chatbot does not know your offer, prices or orders

An off-the-shelf tool replies with generalities, because it has no access to company data and cannot check anything in the systems you already use.

After hours or in peak times nobody can reply in time

The question arrives in the evening or in a busy stretch. The customer waits, and the team only gets back to it when they have a moment.

An assistant only helps when it uses your company data and can carry out a specific task. Show me which questions come back most often.

Tell me about your process

What I can do

The assistant uses company data and carries out specific tasks

Not a generic chatbot on the website. I build an assistant that answers from your offer, documents and systems, and can carry out an agreed step.

Answers based on documents, your offer and FAQs

It answers the way the company does — from your materials, not from generic phrases.

Questions about the offer, prices and scope

FAQs, procedures and exceptions

Documents the team currently hunts for by hand

Checking orders, availability and customer data

It checks a specific case in the systems you already use, instead of sending people to the next person.

Order or delivery status

Whether a date is available

Basic customer details

Creating tickets, bookings and updates

After we agree the rules, it does not only reply — it carries out a simple step in the process.

A new ticket in the system

A booking or a status change

An update that someone currently clicks by hand

Technologies I work with

Mastra
PHP
Laravel
React
TypeScript
REST APIs and webhooks
PostgreSQL

How we work together

From a problem to a working solution

You do not need a finished specification or a chosen technology. First, we decide what is worth doing and which first step lets the project move forward safely.

01

You describe the problem or process

You show me what is not working today, requires manual work or takes too much of your team's time. I ask about the goal, the systems you use and the important exceptions.

02

We agree on the first stage

When the scope is clear, you receive a work proposal and an initial estimate. When there are more unknowns, we start with a short paid analysis, specification or prototype.

03

I deliver the solution in stages

Each stage has an agreed scope and result. I show progress regularly, test the important cases and discuss changes before they increase the time or cost of the work.

04

We launch and check the result

I launch the solution, hand over the information you need and we check whether it actually improves the process. Then we can close the project or plan the next stage or ongoing support.