$ cat README.md▋
The X stands for whichever generation comes next
X-GenAI builds AI automation that keeps working long after the demo ends.
GenAI didn't stop at the first wave of chatbots, and it won't stop at whatever's newest right now either. The X in the name is deliberate: a stand-in for whichever generation comes next, 1st, 2nd, nth, whatever a workflow needs. X-GenAI builds custom AI agents and workflow automation engineered to outlast the model they shipped on, so an upgrade upstream never turns into a rebuild downstream.
In practice that's three things: building automation that holds up in production, consulting on where AI is genuinely worth the investment before a client spends on it, and writing openly about what's working, what isn't, and what's worth learning next. That's what the blog and newsletter are for, not another sales channel.
How X-GenAI works
Consulting before building
Every engagement starts with figuring out where AI creates real value for your business, not just what could technically be automated. If the honest answer is "don't build this," that's what gets said.
Education, built in
The blog and newsletter track what's changing in AI: real advancements, not hype, because a client who understands the tech gets more out of it than one just handed a black box.
Built for business growth
Automation gets judged by hours saved, revenue unlocked, or headcount that didn't need to grow, not by how impressive it looks in a demo.
Engineering discipline
Every build ships with logging, eval sets, and a review gate before it goes live. The gap between a demo and something production-ready is usually just whether anyone's still watching after launch. X-GenAI watches.
Pragmatic tooling
Real code where it matters, low-code where it's genuinely faster: whichever gets a better result, not whichever's trendier.
Compliance by default
GDPR isn't an afterthought. Most clients are based somewhere it isn't optional, so it's built in from the first workflow diagram.
One point of accountability
Whoever scopes your project stays on it start to finish. No handoff to an account manager relaying to an engineer you've never spoken to.
The mission
01
Responsible by default
Refusing to ship something just because the technology makes it possible. A decision about someone's academic record, a medical note, a financial approval: those get more review, not less, the more automatable they look.
02
Climate-conscious by default
The smallest model that reliably gets the job done, not the biggest one that looks impressive in a demo. Better for the client's bill, better for the planet running the servers behind it.
Review gate · every build
Model choice · default
X-GenAI holds every build to the standard laid out above. For proof before committing real budget, that's exactly what the Discovery Audit is for: a focused, time-boxed review, credited back against the final bill if the project goes ahead.
Based in Amsterdam. Working with clients anywhere.
Where things stand
- ▸ Amsterdam-based, working with clients anywhere, remote by default
- ▸ Every build ships with an eval set, logging, and a review gate: the default, not an add-on
- ▸ GDPR-aware from the first workflow diagram, not retrofitted afterward
Curious whether your workflow is automatable?
15 minutes, no pitch deck, no sales rep in the way. Just whoever would actually build it.