Senior full-stack engineer · Ruby on Rails

Plan A was to be a professional hockey player my whole life. I ended up somewhere else.

Senior full-stack engineer. I build pricing engines, scrapers and LLM pipelines in Ruby on Rails.

6+ years in production, working remote from Czechia. Before that: ice, then cars, then code.

Matěj in an HC Zubr Přerov jersey, number 86 — #86, HC Zubr Přerov, January 2016.

§ 1 — What happened

The short version.

Pro hockey until 28. Two years selling cars at a Škoda dealership. Then I taught myself to code, and since 2019 I've been building the core of Disivo, a dynamic pricing SaaS.

I. — until 28
Professional hockey

Plan A. The only plan, for most of my life.
II. — 2017–2019
Selling Škodas

Two years at a dealership. Learned how to deal with people.
III. — 2019 → now
Software engineer

The core of Disivo. Plus my own iOS apps on the side.

§ 2 — Somewhere else

Where I ended up.

Problem, what I built, what changed. Production systems at Disivo.

Pricing engine

Rails · PostgreSQL · Elasticsearch · Redis

Built and maintain the core of Disivo's dynamic pricing platform: the engine that turns competitor data into prices for online stores.

~⅓of Czech e-commerce

Scraper without a browser

HTTP + parsing · proxies

Headless rendering was slow and expensive. I built a scraper API on plain HTTP requests and parsing, running in production next to the old system, so a new source goes live fast.

0headless browsers

Zora

LLM extraction · RAG

A production LLM extraction pipeline, not a demo. Next-generation scraping of a million products per day.

1Mproducts per day

§ 3 — What I do

What I'm actually good at.

I like building web apps that stay clean as they grow, and putting LLMs to work on real business problems. Six years in production taught me to own things end to end, from the scraper that collects the data to the price that goes live.

full-stack: Web apps built to last; Ruby on Rails, Hotwire, PostgreSQL, Elasticsearch, Redis. Boring tech, used well.
genai & llms: AI that ships; Extraction pipelines, RAG and LLM features that run in production behind a queue, not in a notebook.
data at scale: Messy data, tamed; Large-scale scraping and pricing data for roughly a third of Czech e-commerce.
versatility: Quick to pivot; Hockey, then sales, then software. After that, picking up a new stack is the easy part.

§ 4 — On my own time

Three apps, one retired.

iOS apps I designed, built and shipped alone.

A — iOSApp Store ↗ Skycatching Collect the real stars above you tonight, based on where you stand and when. Live on the App Store. B — iOShomeborn.app ↗ Homeborn Documenting a house build from the foundations up. Made while building my own. Live on the App Store. C — iOSretired Filvy A family document vault. Designed, built and shipped alone, then stopped: Apple made the idea native with OCR search across scanned documents and photos. knowing when to stop counts too.

§ 5 — Contact

Let's talk.

LinkedIn ↗

stack: Rails, Hotwire, PostgreSQL, Elasticsearch, Redis
also: Scraping at scale, LLM / RAG, iOS
based: Vysočina, CZ · remote, CET
speaks: Czech, English