/tech

software and hardware development is changing abruptly in the times of ai. modern development is less about one rigid stack per developer, but much more about orchestrating various machine learning models and their context, with not-so-rigid stack. this allows for bigger flexibility of a singular developer that can now prototype-to-product faster

what i do

automation & workflows designing and building automated pipelines that connect your data and tools and eliminate repetitive tasks. n8n, custom scripts, API integrations.

ai integration connecting large language models to your data and tools. MCP servers, RAG systems, AI-enhanced applications that actually solve problems.

backend development python-first, with experience in c++ and modern web frameworks. clean code, documented, maintainable.

infrastructure docker, kubernetes, linux servers. i run my own homelab and understand what it takes to keep systems running.

how i work

i ask questions first - understanding your problem matters more than jumping to solutions.

i prefer simple over clever - code that works and can be easily maintained beats elegant abstractions nobody understands.

i document everything (e.g. curate ai-gen docs)

experience

freelance developer / 2025 - present / custom ai solutions, automation workflows, full-stack development for clients worldwide.

junior software developer / Seznam a.s. / 2022 - 2025 / backend development in c++ and python. docker, kubernetes, openstack. linux systems.

www apps developer course / 2021 - 2022 / html/css, javascript, c# .net, ms sql.

ai & ml

openai api   → gpt integration
anthropic    → claude integration
claude code  → local llm interface
open webui   → local llm interface
llama.cpp    → local models

languages

python       ████████████████████ daily driver
javascript   ████████████░░░░░░░░ when needed
typescript   ████████░░░░░░░░░░░░ growing
bash         ████████████░░░░░░░░ glue & automation

frameworks & libraries

fastapi      → apis, backends
langchain    → llm orchestration  
langgraph    → ai agents
svelte       → frontend (personal choice)
react        → frontend (when required)

infra

linux        → servers, daily driver
docker       → containerization, model running
docker swarm → orchestration
traefik      → reverse proxy

databases

postgresql   → primary choice
redis        → caching, queues
mongodb      → when documents make sense
timescaledb  → time-series data

automation

n8n          → workflow automation
mcp servers  → llm tool integration
custom apis  → whatever connects things

other tools

git          → version control
uv           → python package management
cursor       → editor
warp         → terminal 
raycast      → better spotlight
homebrew     → macos package management
obsidian     → notes