01
Use-case analysis
We work out what your assistant should do, which questions come in and where humans take over.
An AI chatbot that knows what it's talking about
Assistants that actually know your business: trained on your content, with clear guardrails and a clean handover to your team. While you read this, our bots are answering real customer questions somewhere.
At a glance
1 knowledge base
Answers come from your content, not from the open internet.
100% your rules
Guardrails define what the bot says and when it hands over to humans.
Web & WhatsApp
Your assistant talks where your customers are, including messengers.
24/7 in production
Our assistants talk to real customers around the clock.
Not a cookie-cutter bot
An assistant that knows your business
Trained on your content, connected to your systems and limited to what it is allowed to say.
Where chatbot projects fail
A chat widget is added quickly. Things get difficult once the bot needs to be reliable, legally safe and useful.
01
Hallucinated answers
A bot that makes things up costs trust. Without a knowledge base and guardrails, AI becomes a liability.
02
Widgets without connections
Off-the-shelf widgets know neither your products nor your CRM. The bot chats away while your team keeps typing.
03
Legally sensitive
Health, finance, advice: in regulated industries a bot must not say many things. Standard tools don't know that.
04
Leads left behind
The bot answers questions, but nobody hears about it. Without handover to people and systems, the contact evaporates.
What your AI assistant can do
Everything here comes from assistants that talk to customers in production today. Combine what your use case needs.
01
Your own knowledge base
Your bot answers from product data, help articles and documents, kept current and limited to your topics.
02
Guardrails & compliance
Fixed rules for what may be said. Including regulated industries such as health or finance.
03
Page context
The assistant knows which product or page your customer is looking at and answers accordingly.
04
Lead handover & routing
Conversations become contacts: the bot qualifies enquiries and hands them to the right person.
05
WhatsApp & messengers
Your assistant also replies where your customers already write.
06
Connected to your systems
CRM, shop, calendar or database: the bot reads and writes where your data lives.
07
Multilingual
Your assistant speaks your customers' languages and switches mid-conversation.
08
GDPR & EU operation
On request with EU hosting and European models, including deletion and disclosure processes.
09
Insights & learning
You see what gets asked and where the bot needs sharpening. That's how your assistant improves.
How you get your AI assistant
Four steps from analysis to a running assistant.
02
Concept & fixed price
Building on that, we plan concretely: knowledge sources, channels, guardrails and a fixed-price quote that holds.
03
Training & integration
We build the bot, feed the knowledge base, set the guardrails and connect your systems.
04
Launch & fine-tuning
Go-live with insights: we look at what gets asked and sharpen answers and rules.
Ready?
Let's talk about your assistant
In a free consultation we'll work out what a chatbot can take over for you and what it had better not do.
Frequently asked questions
An assistant with its own knowledge base for your website usually starts at around €6,000, and most projects land between €8,000 and €25,000. WhatsApp connection, CRM integration and lead routing extend the range upwards. The deciding factors are knowledge sources, channels and how strict the guardrails need to be. After the free consultation you receive a fixed-price quote, and if a ready-made tool covers your case, we will tell you honestly.
As soon as the bot should really know your business: your product data, connections to CRM or shop, your brand voice, legal boundaries or lead handover to your team. For a simple FAQ list, a builder is often enough. For anything that touches revenue or liability, custom work pays off.
That's exactly what we prevent, on three levels: the bot only answers from your knowledge base, guardrails limit its topics, and when it doesn't know something it says so and hands over to a human. Before go-live we deliberately test with tricky questions.
The leading models from Anthropic and OpenAI, and depending on the case European models with EU hosting for sensitive data. The choice follows your use case, your privacy requirements and the cost per conversation. Switching models later remains possible.
Yes. We have assistants in production that talk to customers via WhatsApp, including conversation history and handover to the team. Web widgets and other messengers are possible too. The right channel depends on where your customers write.
Yes, if it's built correctly: data-processing agreements with the model providers, EU hosting on request, data minimisation in prompts and clear deletion processes. Your customer data does not flow into training other people's models.
Yes. The assistant can read and write in your CRM, shop, calendar or database: quote availability, check order status, create contacts, suggest appointments. Exactly this connection separates a useful assistant from a chat toy.
It says so honestly and hands over cleanly: to the right team member, via ticket or with a callback request, including the conversation context. Nobody has to explain their issue twice. You define the handover rules.
A focused assistant with a knowledge base goes live in a few weeks. With messenger channels, system connections and lead routing it takes longer. After the analysis you get a concrete timeline, and we deliver in stages so you can test early.
Get in Touch
Tell us which questions your team answers over and over every day. In a quick call we'll work out how a chatbot can take them on reliably, trained on your own content.