The problem with German software startups isn’t “just” bureaucracy, taxes, etc. etc. Let me give you another major problem and example: access to open data and APIs.
ALL successful US startups have a developer page where you can register and use their APIs. In 90% of cases, free of charge, so you can build your own software on top of them or simply connect your existing software with their services.
And they benefit from this too. If their customers can easily connect their tools with your ERP, CRM, or any other software, their own product becomes more useful and attractive as well.
But no, here in Germany we of course have to make that paid AND DEFINITELY NOT PUBLIC.
DATEV wants several thousand euros in fees because they want to “review” MY WORK, THE CODE I WRITE AND THE TIME I SPEND ON IT, decide whether it’s okay, and only then give me access to the API.
And if it’s not okay, it’s not even your problem anyway. My fckng software simply won’t work and you’ll get, at most, an error message.
Just wanna publish one of my public projects for the developer community.
Built this over the weekend because I had this problem myself and I’m pretty sure a lot of developers still using VS Code have it too.
Most developers are using Claude and Codex now. The problem starts when you have one huge chat, all the context is there, everything is working perfectly, and then...
Your tokens are gone.
Now you wanna continue in Codex instead of Claude, or the other way around, but you can’t just move the whole conversation with all the context and continue working.
So I built exactly that.
You can take your current chat, create a handoff, move it to the other agent and just continue from where you stopped.
Claude to Codex.
Codex to Claude.
No starting from zero again. No manually explaining your whole codebase, what you changed, what you tried and what still needs to be done.
I also added it as a VS Code plugin, so you can use it directly while coding.
It’s public on GitHub.
Just check it out, run it and tell me what you think.
github.com/ibodc/agent-switc…
Attached a small video showing how it works 👇
Ibo Dc retweeted
Nobody talks about how hard it is to spend your whole life writing code and then become the person who has to sell.
I never thought sales would be part of my job. For the first three months I was bad at it, and I lost deals I should have won.
The mistake I kept repeating was treating a sales call like a demo. I would show what we built, walk through the parts I was proud of, and wait for the prospect to be impressed. They usually weren't, because I was answering questions they hadn't asked yet.
Four things that helped:
Give yourself permission to be bad at it for a while. You spent years getting good at engineering, and nobody expected you to be good at that on day one either. Sales works the same way, except the feedback loop is slower and more expensive.
Book demo calls with other companies, including tools you have no intention of buying, and pay attention to how the person on the other side qualifies you. You'll notice they ask far more questions than they answer, and that almost all of those questions come before any product shows up on the screen.
Rehearse before it counts. Mock calls with a colleague, or even saying your pitch out loud alone, will expose the parts of your story that don't hold together. Finding those gaps in practice is much cheaper than finding them in front of a prospect.
Listen longer than feels comfortable. Prospects describe their problem in their own words, and those words are almost always better than the ones you wrote yourself. I now reuse the exact phrasing customers give me when I write outreach.
Engineering taught me to build the answer before anyone asks for it. Sales asks you to do the opposite, and unlearning that habit took me longer than learning anything new.
Has a point... some things are just much, much easier to do manually.
I think building in public will create a lot of the next generation of billionaires, or at least very successful entrepreneurs.
Over the next few days, I’m going to share my next big thing.
Applied to @ycombinator again.
This time with something realllyyy different—and genuinely exciting. Let’s see what happens.
Once we get accepted, I’ll publish the video we made with @atrlergn 👀
I wanna ask @peterthiel if he really still thinks AI will take over humans not before the 22nd century... like he suggested in Zero to One.
Ibo Dc retweeted
🚕 Do you run a taxi, chauffeur, or transfer company and still dispatch your rides manually?
With Doards, you only need to enter the pickup location, destination, and date. Then click “AI Dispatch”, and our intelligent AI automatically finds the best available driver and vehicle.
Ibo Dc retweeted
🚗 Smarter fleet insights, without additional hardware
We’re excited to announce our collaboration with ECHOES Solutions.
Through this integration, Doards can now access valuable vehicle data directly within the platform, without installing any extra hardware or tracking device.
Apple sitting on 2.5 billion devices acting like they run the AI game while literally begging Google and OpenAI for models. Thats not innovation thats integration. Look at what actually made them big. They built Swift. They built their own silicon. The whole iOS ecosystem from scratch. That was real. Now compare that energy to slapping Gemini on Siri and calling it a strategy. You want to win long term you build your own stack. Your own infra. Your own models. Meta did it with PHP. Apple did it with hardware. The moment you depend on someone elses stack you lose leverage. Just look at all the startups running on Vercel. It works until traffic explodes, costs explode, and suddenly your entire business depends on a platform you don’t control.
Germany wants to be a tech hub but try reaching out to any mid size company there about integrating with their systems. You will get silence. Absolute ghost mode. Not even a rejection, just nothing. The whole ecosystem sits on closed doors and unread inboxes. Meanwhile they complain about falling behind in digitalization like it is some mystery. Brother the talent is there, the builders are there, but the culture of gatekeeping kills everything before it starts. Open source exists but the mentality to actually collaborate does not. You cannot build a digital future while hoarding every endpoint like state secrets.
Sorry there is euroleague.. worst ticket system ever
Ticketmaster is the final boss of broken systems that nobody can kill. 30k people in a digital queue. For concert tickets. In 2026. And then you get blocked without a reason and can't unblock your account..
The entire AI industry is rewriting how humans work and we still cant buy Harry Styles tickets without a PhD in queue psychology. Everyone building the next chatbot while billion dollar monopolies run on infrastructure from 2008. The unsexy problems stay unsexy because disrupting them requires lawyers not engineers.
Doards is going international 🌍
With our latest update, we are officially expanding beyond Germany. Our first clients from the UK and the UAE are currently onboarding.
For now, we are focusing on English-speaking markets. The goal is simple: build it properly, make it stable, and ensure everything works flawlessly before expanding further.
International growth is not about moving fast — it’s about building something that works across markets, compliance systems, and operational realities.
This is just the beginning.
#Doards #SaaS #MobilityTech #ERP #Startup #InternationalGrowth
Years ago I started my own company and while working closely with customers — especially in the mobility industry like taxi and car rental — I realized something big.
There is a huge gap in ERP.
Whenever we talk about ERP, the first name that comes to mind is SAP. It’s probably one of the most robust systems in the world. Extremely powerful. But also extremely complex. Hard to learn. Hard to sell. Expensive to buy. And if you really want to customize it deeply? That easily costs millions.
For large enterprises that might be fine.
But what about small and mid-sized businesses? Especially in mobility? Companies that don’t have that kind of budget but still need structured processes, compliance, automation, reporting — the full professional setup?
That’s why I founded DOARDS.
The original purpose was very clear:
Customization.
Not just another default SaaS where you adapt your business to the software.
But software that adapts to your business.
The idea was almost like Shopify or WooCommerce — but for real operational businesses. A system where, with some programming and design knowledge, you can build your own ERP logic on top of a secure, compliant foundation. Use default modules if you want. Or adjust everything.
Then AI arrived.
And the whole software world started changing again.
So we built our first AI model: DISI.
It’s not fully coding everything yet — but it already does something powerful. It reads WhatsApp messages from customers, understands booking intentions, structures the data correctly, creates bookings directly in the database, triggers scripts, and replies automatically.
In other words: it is already operating inside the ERP.
This is just the beginning.
The next model we are building will allow users to actually create logic, use our widgets and functions, and customize their ERP system through AI assistance. Imagine combining AI + modular ERP + real backend access.
And yes — we are doing this with PHP.
Because PHP gives you freedom. It runs the web. It connects easily to HTML (which is literally the whole browser). And for customization-heavy systems, that flexibility matters.
The goal was always the same:
Give businesses control.
AI is just the next step in that mission.
Stay tuned.