AI is cool ‘n all but connecting systems is what brings real value.

Many tech startups fail because they’re a solution to a problem that doesn’t exist. AI won’t help them. Talking to potential users could.
We’re a small software company based in Amsterdam. We build and maintain some pretty complex projects. And, crucially, because we also do support, what we also do is talk to our end users.
But the other thing that is particularly helpful is that we initiate and participate in marketing campaigns and so we also get to talk to leads.
Let me give an example. One of our projects is an inspection tool. When you hear the word “tool” you might think of a hammer or such. But this is a very complex platform with hundreds if not thousands of features.
Talking to a lead one day, we have a friendly conversation about our platform and the one they are currently using. We demo ours, they go, "There’s nothing in yours that we don’t already have".
Ok...
But just as we’re closing off, one of the participants says: “if there’s one thing I’m missing, it’s the ability to bring up a case file directly from our app”.
Bam.
Exactly what we're working on. We’d even built a proof of concept that could connect to other systems to do just that.
Huh? Isn’t that really easy? Rig up an API and you’re done.
Thing is though, it takes two to tango. To bring up a case file we’d need to connect to their business system. And we’ve yet to convince the other that connecting systems brings great value.
When connecting systems is first task is to bring minds together. The tech comes after.
Our tool is generally used by organisations that either enforce government regulations or must adhere to them. When enforcing, the process starts with a case file in the business system of a governmental entity, mostly a municipality.
But municipalities can be using any number of business systems, some legacy, given the by now quite long history of digital technology. They struggle to keep coherence within their own organisation, let alone with outside systems like ours.
The vendors that supply these systems — municipalities in the Netherlands are responsible for their own IT — are therefore quite busy with maintaining and supporting their systems. Some of these companies are not much bigger than we are, so just don’t have the bandwidth to embark on large scale integration projects.
There are others, however, that have understood the opportunity early on, and already support connectivity between their business system and a designated inspection platform. Our competitors.
Companies that sell integrated systems have a leg up. To compete, we'll have to convince like-minded souls to work with us.
The point being that it’s one thing to listen to your users, to hear what your leads are saying, but to act upon it is another entirely different matter. Integrating systems is technically challenging, especially if they’re legacy. But even harder is to bring a convincing case that would be beneficial for both parties.
And indeed, after an agreement, there’s a lot of technical and legal stuff to wade through.
For the technical stuff, we’ll be using AI coding agents. For the legal stuff, we’ll be using the frameworks initiated by governmental organisations such as the Union of Dutch Municipalities (VNG), which contain rules for data interchange. Helpfully written in computer code. Again, just the thing for our AI.
But the task set upon us humans is to convince the other party, also humans, to embark on this journey with us.
It started as one sentence in a call with a lead. For us, it’s a new phase in our development from a technical crew to become the builders we’ve always aspired to being.
We’ll let AI do the nitty gritty, leaving time for us humans to do the thinking. And the connecting.
78% of VC supported startups fail. So what, they move on to the next. For bootstrapped companies like ours, that's not an option.
We're not in the investors pool. We're a software service company that's been helping SME clients for the best part of 15 years. But so are many other IT companies, such as those who maintain and support business systems for government.
But we're competing with companies that have some big money behind them. Companies that can afford to buy up smaller players and integrate them into their systems. What you end up with are these monoliths that are end-to-end solutions but which come with a lot of strings attached. And allow very little input from end users.
We're a small company, yes, but we're in direct contact with our users, and, as I explained above, also with potential new users. Other small companies have the same abilities. And because, just like us, they're independent, not at the mercy of any VC, they're flexible enough to act upon opportunity.
All we have to do is to convince them of that opportunity.
Header image: Pierre-Auguste Renoir, Luncheon of the Boating Party, 1881.