At ValBit, a customer enquiry does not end up on a pile – it goes to a team of AI agents. Before an enquiry reaches that team at all, it is anonymised and cleaned up by hand: customer data stays out. Then the AI team takes over. One agent accepts the enquiry, the next one plans the solution and estimates the effort, and the third turns all of that into a finished offer in our style and with our templates. At the end, a human reviews it and gives the go-ahead – the legwork in between is done by the agents. This is exactly the way of working you can set up in your own organisation. Let me show you how, using a real example from our own day-to-day work.
The paperwork keeps growing. The team does not.
You know the feeling: an enquiry comes in. You are pleased – and in the same moment you can see all the work ahead of you. Clarify what is needed, think through a solution, estimate the effort honestly, and pour it all into a clean offer that still holds up later.
New people to absorb that work? Not to be found right now. So the work stays with the few who are already at full capacity. Offers go out later than intended. Some never go out at all.
At this point many people think of ChatGPT: a program that answers when you ask it something. Handy – but: you ask, it answers, you carry on. The difference that matters here is this:
A chatbot answers when you ask. An AI agent gets the work done – by your rules.
An agent does not wait for your next question. It has a task, it knows your requirements and it works through them: it reads, plans, writes, files. And several such agents can work together like a well-rehearsed team – each with its own role.
The living example: our own offers
At ValBit, we now draft our offers with exactly such a team of AI agents. Not as a vision of the future, but as part of everyday business. This is how an enquiry moves through our house:
Step 1 – The enquiry arrives. Someone sends an email or calls: “We could do with someone for …”. Instead of that note sitting around, it is logged as a case.
Step 2 – A human anonymises and cleans up the enquiry. And that happens before anything goes to the AI. Names, company, contact details – everything tied to a person is taken out. Only the anonymised, clean task goes to the AI agent team. This matters to me, so let me say it clearly: customer data does not travel into the AI unfiltered at ValBit. Data protection is not an afterthought, it is the first step.
Step 3 – The AI agent team takes over. Now the anonymised enquiry runs through the team, each agent in its own role: the team manager accepts it and opens a clean case. The IT expert works out the best, simplest solution and estimates the effort honestly – rather with a buffer than talked down. The offer expert turns that into a finished offer: in the right layout, with our typeface, our language, clearly worded.
Step 4 – A human reviews and approves. At the end, I read the offer, change what needs changing, and approve it. I will say that openly, too: the agents take the legwork off my hands, not the responsibility. Only reviewed work goes out. Whatever reaches the client, I have seen.
And in your organisation?
The offer process is just one example. The principle – a human prepares and delegates, the agents handle the recurring legwork – suits many tasks: offers, confirmations, research, social media posts, reports.
If you would like to build your own team of AI agents: that is exactly what we show you in our one-day seminar.
We look forward to hearing from you about AI agents.