Almost every company has an idea that keeps getting pushed back. But it is hard to invest in uncertainty. Until you try something, you do not know whether it makes sense or whether it will work.
That is why we will choose three applications with the strongest potential and build a working prototype for them for free - work worth up to EUR 3,300. We do not wait for the deadline: places are filled as we go, and if an application grabs us, we start right away. No lengthy specification, no commitment. The point is to quickly find out whether the idea is worth the next step - and you leave with evidence, not just a feeling.
A website, configurator, calculator, or customer portal that could sell better. A process held together by Excel. An internal tool that would save hours of work. Or an AI agent that would sort inquiries, monitor data, prepare materials, and free people up for higher-value work.
We are not looking for a contest for the “best app”. We are looking for real problems with an owner, impact, and room to grow.
Place 1 — taken
An editorial assistant that finds its own topic, checks the archive to confirm it has not been covered yet, and puts the finished text, graphics included, into the CMS through MCP.
Place 2 — open
We decide as applications come in.
Place 3 — open
What matters is whether the problem has an owner and impact — not what industry it is in.
A few minutes and a few sentences about what is slowing you down or what you would like to try are enough.
At Sinfin, we design and build custom web applications, configurators and pricing calculators, customer zones and client portals, e-commerce, auctions, and publishing platforms. We also build internal systems that replace a spreadsheet, AI agents for one repeated process, and integrations with the systems you already use, over an API.
A few things we have built recently. The same pattern every time: a specific problem, a specific solution.
You sell well, but only at home. We connected AI to an e-shop that translated the entire catalogue into EU languages - keeping technical terms, prices, and shipping right for each market, and preparing each version so Google reads it as a local-country website. No translators, no per-page translation billing, and automatic.
Customers cannot picture the product. We built an app where customers can use their own photos to arrange pictures on a wall, view them on their wall through mobile Augmented Reality, and get a hanging plan. Less hesitation, larger carts.
You want to sell live and remotely. We launched live auctions with a near-zero-latency stream from the room and real-time bidding - and handled hundreds of people at once when money was on the line.
Someone in your company spends whole days matching payments by hand. We programmed a process that downloads bank and payment gateway transactions, uses AI to match invoices, calculates fees, and prepares documents for accounting. A person just approves. Tested on a combination of e-shop, bank, Flexibee, and wflow, and it can find forgotten invoices and receipts in email and on disk.
You collect ideas for a company blog, but you do not have time to write. We built an agent that turns saved posts into a verified article in perfect Czech, creates images, and after your approval publishes it in twelve more languages - including an audio version.
You sell one-off, even though customers would gladly come back. We prepared a subscription for a coffee roastery: the customer sets once how often they want coffee delivered and then does not have to do anything else. One-off purchases become recurring revenue.
Every time, it starts with an ordinary problem or an idea that was put aside - and ends with something that saves time, opens markets, or earns money. Your “what if” can be next.
Write a few sentences about what is not working. Places are filled as we go; applications are open until 31 August.
