Pieter Deramoudt · freelance · Ghent
Hi, I'm Pieter.
I help SMEs automate repetitive work.
At the car-sharing cooperative Dégage I automated my own job in the space of a year, using AI and other tools. A person always checks the result. Now I help other organisations do the same, preferably with the systems they already use.
Does this sound familiar?
“We retype the same things every week.”
Then we can look at whether that work can be automated. We could start, for example, with one task that comes up regularly.
“We are drowning in email and don't know what to handle first.”
First we look at how you handle your email today. After that we can work out what can happen automatically and where someone should still take a look.
“Everyone is talking about AI. Where do we start?”
I would start with a process you know well and that involves a lot of repetitive work. That way we can find out together whether AI can be of use there.
“Our system is old. Which new one suits us?”
I can help you compare the options and make a choice. I can also support you during the switch and the migration. I don't sell software myself.
“My team has to believe in it too.”
I understand. That's why I involve the people who will have to work with it from the very start. I also explain and train, so that they can carry on with it themselves afterwards.
Is your situation not on the list? Feel free to describe it and tell me what is going on.
What I can help with.
Automating repetitive work
Emails, quotes, invoices and scheduling. We look at where automation is useful, with AI or with a simpler solution.
Teaching your team to work with AI
Together we look at how AI can help in the daily work. With examples from your own organisation.
Choosing a new system
I help you compare different solutions and see which system fits the way you work.
Guiding an ERP project and migration
I can support you with the analysis, the testing and the follow-up of an ERP project. I worked as an ERP consultant in SAP and Business Central for more than five years.
Do you work at a consultancy or a larger organisation? There is more on separate pages: for consultancies and for larger organisations.
What working together looks like.
- 01
A call or a workshop
We can start with a half-hour call, without obligation. Or we book a two-hour workshop with the team straight away, at a fixed price.
- 02
One process
Together we choose what we want to tackle and agree a budget in advance. The rest of the work is billed by the hour, with an agreed ceiling.
- 03
You take over
I document what we have made and show you how it works. That way you can carry on with it yourselves.
A few ground rules.
A person checks the result
AI can make suggestions and prepare work. Someone on your side checks the result.
Your systems first
We first look at what is already there and what is possible within it.
Clear terms
We agree beforehand what we tackle and how much time and budget is set aside for it.
Sharing knowledge
I explain how everything works and make sure your team can work with it.
A short introduction.
I'm Pieter Deramoudt from Ghent, Belgium. I worked as an ERP consultant for more than five years, first in SAP at Capgemini and then in Business Central at ESC.
After that I became IT manager at Dégage, a car-sharing cooperative in Ghent. A small organisation with many different processes and, at times, quite a bit of chaos. In the space of a year I automated my own job there with AI, always with a person checking the result.
I like efficiency. I can be a bit lazy at times, but that may also be a good reason to prefer automating repetitive work.
Since November 2026 I have been running Gaston Consulting as a side business. Gaston is my middle name.
I don't have any client testimonials yet. I do have plenty of experience with files that badly needed some help.
A testimonial, from an Excel fileFor years I was the single source of truth. Now I'm an export that arrives automatically every Monday. I sleep much better.
Shall we have an introductory chat?
A few sentences about your organisation or the question you are dealing with are enough. I reply myself, and then we see whether a call would be useful. And if nothing comes of it, that is perfectly fine too.
Would you rather talk on the phone? Email me your number and I will call you back.
Not sure what to write?
You can ask your own AI to help you structure your question. For example, paste this into ChatGPT, Claude, Gemini or Copilot and send me the result.
I am planning a half-hour call with Pieter from Gaston Consulting. He helps SMEs automate repetitive work with AI, guides the choice of a new system and the migration, and teaches the team to work with it. What do you know about me and my work from our earlier conversations? Put it on one page for him: 1. My company: what we do, how big we are, which software we use, and whether a switch is coming up. 2. The work that recurs every week and takes a lot of time: who does it, how often, in which system. 3. Where it goes wrong: what fails, what gets left undone, what depends on one person. 4. What I have already tried with AI, and what worked or did not. For each point, say whether you know it from our conversations or whether you are inferring it. Do not make anything up. If you know little about me, first ask me three short questions at most. End with three questions I can ask Pieter. Leave out client names, passwords and confidential figures.
Do we already have an appointment?
You can ask your own AI to set out the facts beforehand. You don't have to. For example, paste this into ChatGPT, Claude, Gemini or Copilot and bring the result to our call.
I will soon have a first call with Pieter from Gaston Consulting. He helps SMEs automate repetitive work with AI, guides the choice of a new system and the migration, and teaches the team to work with it. Help me prepare that call. Put on one page the facts you know about my work from our earlier conversations: 1. The systems we work with: which software, what for, and who uses it. 2. The work that recurs every week: who does it, how often, roughly how long it takes. 3. Where it goes wrong: what fails, what gets left undone, what depends on one person. 4. What we have already tried to solve that, and what came of it. Give facts only. Do not propose a solution, a plan or software: Pieter and I will look at that together. For each point, say whether you know it from our conversations or whether you are inferring it. Do not make anything up. If you know little about me, first ask me three short questions at most. End with what you do not know and what I should check with a colleague. Leave out client names, passwords and confidential figures.