I build practical AI-assisted tools, automation workflows, and small product prototypes that solve real problems in my own work and daily life.
My projects usually sit at the intersection of:
- AI-assisted production pipelines
- Browser and data automation
- Local-first dashboards and personal operations tools
- Media workflows for music, video, and publishing
- Lightweight web apps that avoid unnecessary infrastructure
A standalone Python/FastAPI service that forecasts PV production, household load, and EV-charging surplus from InfluxDB telemetry and Open-Meteo weather data.
Focus: energy forecasting, local telemetry, ML pipelines, public-safe extraction from private infrastructure.
A Next.js dashboard for the FIFA World Cup 2026 with schedule, standings, live-style sections, and forecast probabilities.
Focus: dashboard UX, TypeScript, resilient data fallbacks, forecast communication.
A static promo site for a family-friendly animated story project, including character artwork, audio chapters, merchandise previews, and a playable browser game teaser.
Focus: static publishing, visual storytelling, lightweight game embedding, public asset curation.
A small Python automation pipeline that discovers, de-duplicates, redeems, and reports mobile-game gift codes through Playwright and Telegram notifications.
Focus: browser automation, reliable state handling, headless workflows, personal productivity.
A single-file browser tool for tracking EV inspection intervals, overdue maintenance items, and estimated service costs locally in the browser.
Focus: local-first UX, no-backend tools, practical decision support.
Some of my strongest projects are private because they connect to local infrastructure, accounts, production assets, or personal knowledge systems. They include:
- An AI music factory for brief-to-song-to-video workflows
- A smart-home hub for local device control and agent-assisted operations
- A family command center for documents, tasks, calendar context, and local AI assistance
- Media production pipelines for short-form video and YouTube publishing
When a project becomes useful beyond my own environment, I extract a sanitized public version or publish the reusable lessons.
- Ship small working systems before over-abstracting.
- Keep sensitive data local and out of public repositories.
- Prefer clear README files, screenshots, and reproducible setup notes.
- Document lessons learned so a project is useful after the first build.
- Treat AI as a production amplifier, not as an excuse for messy output.
I am actively curating this profile: archiving old experiments, improving READMEs, and turning the best private work into clean public showcases where it makes sense.