Knowledge & data
AI knowledge system
Every business has knowledge that exists in only one person's head. How to price the tricky job. Why one client needs special treatment. What matters when someone new starts. When that person is on holiday, the question goes unanswered. And even the knowledge that is written down somewhere costs time every day: searching, asking around, waiting for a reply.
You'll get …
- 1
Knowledge that isn't stuck in one head
Standards, processes and experience are written down and easy to find – even when the person who knows them is on holiday.
- 2
Look it up, don't interrupt
Staff find standards and answers in your own knowledge base rather than interrupting colleagues.
- 3
A system that grows with you
At first the AI only reads, then it writes drafts and minutes for sign-off, and later it takes on recurring routines.
01Intro
A knowledge system faces inward – customers never see it. It's where your notes, projects, standards and processes live, organised so that AI can search them, summarise, take notes and run defined routines. What sets it apart from a plain knowledge base: you don't just look things up here, you get work done. I work this way myself every day – much of my day-to-day business now runs through a system like this, with AI as a colleague inside it.
Worth knowing: I won't promise a "second brain in one go" – start too big and it falls apart on upkeep. Whether your business is ready for a system like this or Tier 1 is enough for now isn't a matter of guesswork: a small AI audit beforehand shows whether the build will pay off – check first, then build. Sensitive data (clients, patients, staff) stays out or gets its own data-protection concept. The AI never writes to business-critical areas unchecked. And the system only stays alive if the team uses it – which is why "bringing the team on board" is a required part of the project, not an optional extra.
02How it works
- 01
Take stock: where does your knowledge live?
In heads, PDFs, email threads, specialist software, sticky notes on the monitor. Also: which processes keep repeating, who works how, and where does knowledge get lost when someone leaves?
- 02
Build the structure – small and documented
A clear way of organising files and notes, with simple written rules – the foundation for everything that follows.
- 03
Gather and sort what you know
Existing material moves in – curated, not copied. If there's a large backlog of old documents, Intelligent digitisation does the groundwork.
- 04
Connect the AI: read first, write later
At first the AI may only answer questions, find things and summarise. Once that works reliably, writing with sign-off follows: drafts, minutes, entries. Permissions and limits are set out in a written work instruction.
- 05
Add routines – one at a time
Recurring reports, logs, checklists, reminders: the AI takes on defined tasks within clear limits. Not all at once – each new routine only once the previous one is running smoothly.
03Stages
Every project starts with Tier 1, the team knowledge base.
Starter: the team knowledge base
One clearly defined area of knowledge is organised so you can ask it questions – standards, processes, answers. Staff look things up instead of interrupting.
Next step: the AI joins in
Summaries, drafts, minutes, log upkeep – the AI writes in the system alongside you, and every change needs sign-off.
Full build: the agentic system
Defined routines run on their own – reports, reminders, recurring tasks. The knowledge system becomes a working system. To get there, it needs deep access to your data and processes. That's why Tier 3 only comes after a baseline assessment, once Tier 2 has proved itself for a while, and runs as a project I support throughout.
04Demo
The knowledge system of the fictional “Brandl Carpentry” – on the left the order in which the company's knowledge lives, on the right three everyday moments with it. Click through and watch where it lights up in the tree.
The architecture behind it has two layers: at the bottom the vault – structured folders and notes that belong to the business. Above it the AI layer with documented rules: it reads only at first, then writes with approval, and sensitive areas stay locked. Calendar and tools dock on, routines run within clear limits.
The proof of this model: this website came about this way – I work in such a system myself every day, with AI as an employee inside it.
05Technical implementation
- The knowledge layer belongs to you: a Markdown vault (e.g. Obsidian), Notion or a similar system – your content, no vendor lock-in.
- The AI layer is an assistant with access to the vault (e.g. Claude Code as the agentic layer), working to documented rules: which areas, which permissions, which routines.
- Permissions in stages: read access first, write access only in defined areas and with sign-off, sensitive areas excluded.
- What matters isn't the tool but the combination: clear structure + clear rules + sign-off in stages.
- Upkeep routines are set up from the start – a knowledge system without a maintenance process goes stale like any wiki.
- Typical sources: standards, checklists, text templates, project and meeting notes, onboarding material, lessons learned from projects.
for context · This shows how something like this usually gets built – not how it has to be built for you. What actually fits is worked out in the project: around your existing setup, your team, and what stays easy to maintain.