Quest log

Projects and experience

Long-lived platforms, production integrations and a few tools of my own. Client systems are private, so here is what I owned on each, which constraints mattered and what the work taught me.

Client work

All through a long-term collaboration with DataLite since 2017, first as an employee, for the last several years as an independent contractor.

Main quest

TSM / CETIN

2020 – now

Senior backend and integration engineer, team lead

Backend and integration work on a modern service management platform for a telco operator. I work close to production, on releases, B2B communication with partners, process rules and data compatibility.

Led the 2.2 production rollout
Leading the 2.4 rollout and its release batches
Core developer on version 1.5
Integration flows: ticketing, NIP, change management, B2B, PEW
Incident analysis from logs, metrics and data
Planning, explaining and code review as the main developer for our part of the platform

Stack: Kotlin, Java, Spring, PostgreSQL, Kafka, Kubernetes

Side quest

Nordic web

2024 – now, on demand

Ongoing collaboration on a Django-based web system. I pick up focused changes when they're needed.

Self-service features and web behaviour in Python, Django and templates

Stack: Python, Django, templates

Completed

MNCP

2018 – 2021

Full-stack engineer

Measurement, archives, exports and operational workflows. This is where I moved from delivering features to owning behaviour across backend, frontend and database.

Authentication, export fixes, archive filters and measurement batches
Database scripts and deployment alongside the application code
Tracing bugs through the whole request and data flow

Stack: Kotlin, Spring, Angular, Oracle, PostgreSQL

Completed

ZIS

2017 – 2018

Enterprise Java

An enterprise application for CETIN operations: Java and ZK screens, controller, service and DAO layers, imports, Excel processing and production defects.

Complete business screens from the UI down to persistence
A classic enterprise stack with strict process and data rules

Stack: Java, ZK, ZUL, Maven, SQL

Own projects

Smaller things I build and run myself, mostly together with AI agents. They keep me close to the whole loop: an idea, a usable interface, data, deployment and maintenance.

Parcel pickup map

Every parcel pickup point and box in the Czech Republic on one map, around 40,000 of them, imported daily from carrier feeds and deduplicated.

Next.js, PostgreSQL, Drizzle, Leaflet

Agent toolkit

MCP servers that let AI agents work with real systems: read-only log search over Elasticsearch (working prototype), Jira and a team-chat bridge that sends only after I confirm. Plus a skill library and a knowledge base shared by Claude Code and Codex.

Python, FastAPI, Kotlin, Docker, MCP

Family portal

Self-hosted apps on a small home server: meal plans with real nutrition data, a planner for everything around the house, trips and packing lists.

Next.js, Drizzle ORM, PostgreSQL, Raspberry Pi

Trading bot

A trend-following strategy (EMA crossover, ATR stop) for testing assumptions against market data, not for selling certainty.

Python, Freqtrade

This website

Static, bilingual, cheap to run and easy to change. The source is on GitHub.

Astro, TypeScript, Cloudflare Pages

Before DataLite

Software tester at GMC Software (now Quadient) in Hradec Králové, 2013–2014: use cases, test scenarios and running the team's Jenkins server.

Want to talk about a project?

Long-term or focused contracts, remotely from near Prague.

[email protected]