AI Agents & Assistants
An AI colleague who knows your business.
We build AI assistants and agents on top of your own knowledge and tools. They answer questions, draft replies, process documents and handle routine actions, with guardrails and a human in the loop where it matters.
- Free strategy session
- Fixed-price proposal
- Senior team only
// sound familiar?
Tap everything that hurts.
Be honest. Nobody’s watching.
// tap the problems you recognise
Fix this for me// what you get
Everything you need. Nothing you don’t.
In plain English: AI that knows your documents and tools, does real work for your team and never ends up in the news for the wrong reasons.
AI use-case workshop
Where AI saves real time and money in your company, and where it doesn’t.
Knowledge assistant
Answers from your handbooks, contracts and wiki, with a link to the source.
AI agents
Agents that sort, draft, look things up and update records in your tools, with approval steps.
Document AI
Invoices, forms, CVs and emails turned into clean data.
Guardrails & compliance
EU hosting, access rights, logging and a clear AI policy for your team.
Team training
Hands-on sessions so your people use AI well, safely and every day.
// how it works
Fast, transparent, zero surprises.
Workshop
Use cases ranked by value.
Pilot
One assistant, real users, real data.
Roll out
More teams, more tools.
Improve
Answers, costs and adoption, monitored.
// pricing
Clear prices. Fixed proposals.
Every engagement starts with a free strategy session. After that you get a fixed written proposal, so the number never changes halfway through.
// the fine print, made big
We don’t take every client.
Not because we’re fancy. Because we only take on projects we’re sure we can move, and we put senior people on every one of them. That means a small number of new clients at a time. When we’re full, we’re full.
See if we’re a fit 2-minute application. Honest answer, even if it’s a no.you?
you?
you?
3 seats left this quarter. When they’re gone, applications stay on file.
// faq
AI agents questions, answered.
Which AI models do you use?
We choose per task: models from Anthropic, OpenAI, Google or Mistral through enterprise or EU-hosted services, or open models on your own infrastructure when data must never leave.
Will the AI train on our data?
No. We only use services that contractually don’t train on your data, and your documents stay in your systems.
What if the AI gets something wrong?
Assistants answer with sources, so people can check. Agents ask for approval before anything important, and every action is logged.
Is it GDPR compliant?
We design for the GDPR and the EU AI Act from the start: data minimisation, processing agreements, EU hosting where needed and transparent use.
Can you just train our team?
Yes. An AI workshop or training works on its own, and it’s often the best first step.
// pairs well with
Clients often combine this with
// Q4 2026 intake · 3 spots left
Let’s make AI agents your unfair advantage.
Apply in 2 minutes. If we’re a fit, the strategy session is free and you keep the ideas either way.
* 100% legal. Pinky promise.