59,721
events in the database
Liveupdated every day
Štefan Toporcer · product designer
I've been designing digital products since 2009. Today, with AI, I also build them myself, without a development team: small software tools for boring, bureaucratic problems, for local and global markets.
Engine room
Every project is an engine: a small automated machine for one boring problem. Some are running, some are warming up, others stand and wait their turn. This is the whole engine room.
59,721
events in the database
Liveupdated every day
8engines
Running: Live, working every day.Warming up: Built, or validating demand.Parked: Waiting until another engine proves demand.Stripped for parts: Taken apart, its parts reused.
Workshop operating rules
I want to build my own income from my own products. Today I no longer need a development team for that, and I want to make the most of this opportunity. I do business both locally and globally: some engines solve problems in Slovakia and Czechia, others serve customers anywhere in the world. I start with small engines that tackle a clear problem at low cost. I learn something from each one, even the ones that fail, and work my way up to bigger, more complex projects. These are the rules I use to choose and build every engine.
Hygiene regulations, compliance records, product data stuck in Excel. Nobody enjoys this work, so few people do it properly. Companies are glad to pay for a fix, because the alternative is a fine, a lost distributor or a weekend with a spreadsheet.
I have 10 to 20 hours a week for this. So I only build things that run on their own: the data gets pulled, the document gets assembled, the alert goes out. One more customer must not mean one more hour of my time.
My first projects didn't fail on technology. They failed because nobody got in touch. Now every engine has a number, set in advance, that decides whether it continues or stops. Feed Factory, for example, only continues if at least 8 of the 44 manufacturers in the research confirm the problem.
In industries that still run on PDF forms, a clear questionnaire and a well-typeset document are enough to put a product a class above the competition. This is where my years in design pay off the most.
When an engine produces a legal document, a mistake costs the customer, not me. So I check every mandatory point of a regulation word for word against the official text. I'd rather leave something unsupported than guess. Every document is checked by a person before it is sent, and in Slovakia it is also approved by the regional public health authority.
About

I'm Štefan, most people call me Steve. I've worked in design since 2009: I started with graphic design, moved on to digital products for fintech and most recently designed for the Můj AGEL patient portal in healthcare.
Alongside design, I spent years in sales and tried running my own businesses several times. Some worked for a while, others not at all. B2B sales in Slovakia, Czechia and Cyprus taught me that a great product isn't about code, it's about a solved problem. And that someone has to want it and someone has to sell it.
Every idea ran into the same wall: without a development team, I couldn't build the product. With AI, I now can. That's how the boring engines started: small products I can build myself, meant to earn money.
I do vibe coding: I build products with AI tools like Claude Code and Cursor, from the interface through the database to payments, and I test and verify every output. The goal is clear: work my way up from small tools to my own app or software that solves a bigger problem.
Interfaces, design systems and complete product flows for healthcare, fintech and B2B tools. I know where people get stuck in a form and how to fix it.
B2B sales in three countries, including EdenHaus fleet management. Cold emails, follow-ups, negotiations.
With Claude Code and Cursor I build complete products: data pipelines, Make.com automations, document generation and payments. No team, minimal budget.
Newest first
Decisions, numbers and status changes. Including the ones that didn't work out.
Site
Launching this site: one place for all the engines and the reasons I build them.
Workshop
Consolidation. Out of nine ideas in motion, three keep running and the rest are parked. I have 10 to 20 hours a week for this, not a hundred.
Poriadok.ai
Every mandatory point of the regulation checked word for word against slov-lex.sk. Added every competent local authority: 36 in Slovakia, 14 in Czechia. Czech food service stays unsupported until its exact content is verified.
Feed Factory
Research ready: 44 small manufacturers in the UK, Ireland and Belgium, and an email with no offer, just a question.
Permit Signal
First outreach to companies in Texas. There are 147 companies on the list.
Permit Signal
getpermitsignal.com is live.
Permit Signal
First 30 real dossiers. Finding: the customers will be small operators, not big remediation firms.
Permit Signal
First data load: 58,000 events from 12,000 facilities in Texas, Louisiana and North Carolina.
Workshop
The “Three boring engines” plan: B2B models with near-zero capital that mostly run themselves once set up.
RealAI.sk
The first engine: a real estate listing rewriter, tested end to end from form to finished copy.
Contact
Write to me if you need a product designer who also builds what he designs, if you have a boring problem that should run on its own, or if you'd like to be the first to try one of the engines.
toporcer.stefan@