RS Mobile · Rafał Szewc · Gliwice, Poland
Five projects I built on my own: Turlo, a travel marketplace for web, iOS and Android, and four systems for manufacturing companies in Poland. Each case study covers the problem, what I built, the screens and the stack.
Turlo · Travel marketplace · 2025–2026
Organizers publish a tour by pasting a link to it: an OpenAI vision model reads their page and turns it into a listing in 16 languages, queued for approval. Travellers search in a plain sentence and contact the organizer directly. I built the backend, the web app and the 2.0 iOS and Android apps as the only engineer.
Manufacturing · four case studies
Quoting, material certificates, contract production and shop-floor planning. Each tool started from what the company already worked with: a spreadsheet, a folder of drawings, a supplier’s PDF pack.
Kuźnia Polska · Closed-die steel forgings
The estimator enters the material and the production operations, and the app recalculates the production cost, overheads, margin and the price in EUR and PLN on every change. The second stage is a prototype of a local AI model that reads parameters from customer drawings.
Mostostal Zabrze · Engineering and construction
For a stainless-steel piping job, every line of the bill of materials needed its EN 10204 3.1 certificate from a supplier pack of more than 200 pages. A local vision model reads the certificates; the tool matches them by heat number, writes the certificate number into the Excel sheet and saves each page as a labelled PDF.
Humavit · Dietary supplements
Humavit, a family manufacturer from Mikołów, makes supplements under its clients’ brands and its own. The system calculates the recipe and the margin of each quote, runs raw-material orders and the eight production stages, and pulls invoices from KSeF, Poland’s e-invoicing system, or reads them from PDFs.
Lakotech · Sheet metal · powder coating
Material, thickness, quantity and operations used to live only in DXF file names. One app now turns drawing folders into orders, shows the shop floor today’s and tomorrow’s plan, picks up paint jobs from a WhatsApp group and prepares delivery notes, labels and prices for the invoice.
A few sentences are enough. I’ll reply with questions, or with a suggestion for the first step — usually a proof of concept on your own data.
