I am a data engineer and hobbyist programmer constantly searching for new areas for deep dives.

Data engineering and analysis

My latest project: SELF HOSTED INTEGRATED PLATFORM

Complete Node.js (+ Express) solution for bank-data synchronization, real-time analytics and ML-based transaction categorization. Transactions are fetched via the Nordigen API, enriched, auto-classified by a custom ML model and displayed in an intuitive web dashboard. Users can reassign categories on the fly and trigger model retraining directly from the UI. The app runs on a Raspberry Pi 5 (dev / prod) exposed through Cloudflare, with fully automated CI/CD and declarative cloud resources. Highlights - Multi-environment setup (dev/prod) with zero-downtime deployments - End-to-end test suite and conventional commits enforced in CI - ML pipeline for transaction classification and self-service retraining - Infrastructure as Code for repeatable GCP provisioning - Public URL secured & accelerated by Cloudflare Application tech-stack: - Node.js18 LTS + Express4 - GCP - Cloud Run, BigQuery, Cloud Storage, Looker Studio - Docker Compose + Buildx (multi-arch images cached in GHCR) - Terraform (IaC) - GitHub Actions (lint, test, build, push, deploy) - Raspberry Pi 5 (arm64)

Homelab

Working on my self hosted smarthome