Services · Tools and automation
Less manual work. More from what you already have.
I organise your team’s knowledge and tools, then build apps and automations that take over repetitive work. First, I check what can be done in the environment you already have.
‘I know this from my own work. Someone asks for a quick summary of activities, and the data sits in several spreadsheets and emails.
Someone on the team pulls it together by hand over two days, because the whole system runs on manual work. A month later it all starts over again. Those are exactly the places where I start.’
Your team pays for a dozen or so apps and uses two of them day to day. The same data is retyped from one system into another, and everyone on the team has their own favourite tool. When an idea for AI comes up, the first reflex is to buy yet another licence.
Before you buy anything, I sit down with the team and look at how you really work. I look for what lies unused in the tools you already have (for example Microsoft 365 with Copilot, Google Workspace with Gemini, LLMs). I connect them so that data flows on its own, and I bring in AI where it adds real value. I don’t replace your stack or build anything alongside it. I get more out of it.
What you are left with is a simple map: what we use for what, and how. The team works in the same tools as before, only more efficiently.
A partner campaign is about to start. For six weeks you need to collect submissions, materials and confirmations from dozens of people. Everything ends up in emails and a spreadsheet, and someone spends every day chasing, checking and retyping. An off-the-shelf product does everything except the one thing you need, and a software house quotes a whole quarter.
In a situation like this I build a simple tool for exactly that process: one screen showing who has already replied, what has come in and what is still missing. Reminders go out automatically, and the results are pulled together for you. The first version is up and running within a few days. After the campaign the tool disappears, and what stays behind is well-organised data.
This is how I work across all kinds of processes. I build tools and automations that let the team run activities on its own that it does by hand today. Each one is built for your process, so there are no unnecessary features and nothing that needs adapting. Rollout takes days and doesn’t require any work from your IT team. IT gets a description of what works and how from the very start. They know what is going on, but they don’t hold up the rollout.
Every month the same person compiles a report from five files. Data goes from email into a spreadsheet and from the spreadsheet into a system, and something gets lost at every step. When that person goes on holiday, the report doesn’t get done.
I take a process like this and check which steps follow clear rules and repeat every week or every month. I turn those steps into an automation that runs in the background on your tools (for example Power Automate, Make, Zapier or whatever you already have). Where a human decision is needed, I leave an approval point. I don’t change the way you work, I simply take the manual labour out of it. The report is ready before anyone asks for it. A person approves it instead of compiling it, and the process runs regardless of who happens to be in the office.
I automate all kinds of repetitive processes this way: reports, document workflows, moving data between systems, notifications and summaries. I only automate what follows clear rules, and decisions stay with people. Everything runs on the tools you already have, so there is no extra system to maintain. Your IT team knows from the start what is being built and receives a description of how it works. The team takes over the automation and can keep developing it without me.
A new team member asks where the latest client presentation is. Nobody knows for sure. Some of it is in emails, some in a folder from a year ago, and the rest is in the head of a colleague who is on holiday. The AI you use can’t help either, because it knows nothing about you.
A company knowledge base is one place that holds what there is to know about your company: clients, projects, procedures and agreements. I build it from what you already have, using Andrej Karpathy’s method. AI processes each source once, writes the knowledge into an organised place and keeps it up to date. I don’t move you to a new system. The knowledge base connects to the documents and tools you already use (for example SharePoint, Google Drive, Notion) and speaks your language.
Together we agree on simple rules: what goes into the knowledge base and who keeps it tidy. That way it grows instead of turning into yet another dumping ground. A new team member finds the answer on their own, and AI answers questions about your company, not about companies in general.
I build knowledge bases like this for different areas, for example sales, customer service, onboarding new people and running projects. I always base them on your documents, with a clearly agreed scope of what AI has access to. IT knows where the data is and who can see it. The team receives the knowledge base together with the rules for running it, and carries on developing it on its own.
Let’s talk
First, I look at what your work really looks like: what you put together by hand, where you wait for someone and where the same information is retyped for the third time. Then I build on what you already use: for good, or just for a single promotion, event or campaign. The first working version is ready in a few days, not after a quarter of analysis.