Backend-first developer turning raw data into real-time, production-ready apps.
I build pragmatic systems end-to-end: APIs → streaming → storage → dashboards — and I ship every day.

• Real-time backends (Django/DRF/Channels, FastAPI) with Redis/PostgreSQL/Kafka
• Data/ML pipelines with scikit-learn; learning PyTorch for applied projects
• Clean DevEx: Dockerized stacks, CI/CD, readable docs, and tests

Real-time readings & alerts over WebSockets, token auth, one-click seed/quickstart.

• Highlights: Channels (ASGI/Redis), JWT auth, alert rules, Docker Compose

• Stack: Django · DRF · Channels · Redis · PostgreSQL · Nuxt 3 · Docker

• Repo: Servo-Sens-App

  

Planned features:

• Sensor simulation mode in Docker
• Multi-user auth & role-based access
• Alert rules editor
• Grafana/Prometheus integration

Vehicle telemetry to live map, roles/permissions, alerts, and exports.

• Highlights: Multi-user & roles, Leaflet/Mapbox map, CSV/PDF export
• Stack: Django/FastAPI · PostgreSQL · Kafka · Leaflet/Mapbox · Docker
• Repo: FleetStream

Planned features:

• Real-time aggregation via Spark Structured Streaming
• Geofencing & trip segmentation
• Alert notifications (speeding, idle, low fuel)
• Dashboard with KPIs + map heatmaps

Demo:

From dataset → feature engineering → model selection → /predict API → tiny UI.

• Highlights: Pipelines, 5-fold CV, RMSE/MAE metrics, FastAPI endpoint
• Stack: scikit-learn · Pandas · FastAPI · Pydantic · Docker
• Repo: CarWorth-ML.

Planned features:

• Frontend form → API → instant valuation
• Model retraining pipeline with new data
• Deployment on Hugging Face Spaces
• CSV batch prediction mode

Search by phrase (e.g., “granulate manufacturers Poland”), crawl result pages, and extract emails/phones. Exports CSV/XLSX; ships as a one-file Windows EXE; tiny GUI included.

• Highlights: CLI + GUI, respects robots.txt, follows Contact link (depth=1), CSV/XLSX with hyperlinks
• Stack: Python · httpx · BeautifulSoup · Tkinter · Pandas · openpyxl · SerpAPI · PyInstaller
• Repo: GinioCrawler

“Air-brush” drawing — control the canvas with your hands; clear with a mouth gesture.

• Stack: Python · OpenCV · MediaPipe
• Repo: MultiHandPaint

Try, very interesting!

• Languages: Python · TypeScript · C++ · SQL
• Backend: Django · DRF · Channels (ASGI) · FastAPI · Express.js
• Data/ML: scikit-learn · Pandas · (🔥 learning) PyTorch · Spark · Delta Lake · Kafka · Airflow
• Frontend: Nuxt 3 · Next.js · Vue 3 · Tailwind CSS
• Databases: PostgreSQL · Redis
• DevOps: Docker · GitHub Actions · Prometheus · Grafana

• I prototype quickly, then harden: tests → CI/CD → docs → release.
• I ask “why” before “how”; architecture first, code second.
• AI helps me reason and review. I write the code.

• BSc (Applied Computer Science), 2nd year — Bydgoszcz University of Science & Technology
• Building production-grade open-source projects in public (GitHub commits daily)
• Lifting at the gym 🏋️ and geeking out over cars 🚗

LinkedIn: mateusz-dalke

Email: [email protected]

Open to backend, data engineering, and ML roles — ready for mid-level challenges.