Nil Parra Luna

Networked Computer Systems Administration student (ASIX)

Blanes, Girona

About me

I am in my second year of ASIX, a two-year vocational programme in networked computer systems administration, at Institut Sa Palomera in Blanes. I am interested in network administration, cybersecurity and artificial intelligence.

I like understanding how systems work and finding the cause of a problem before looking for a fix. Outside class I put that into practice in my own projects: a command-line tool published on npm and a machine learning forecasting system that runs unattended on Linux.

After ASIX I plan to take the specialisation course in Artificial Intelligence and Big Data.

My projects

Wooster

Command-line tool and local player to find, archive and watch Twitch VODs with the chat replay in sync.

  • Finds a VOD from an ID, a URL or an approximate start time, without downloading the video
  • Resumable download to a single file, with three download engines
  • Archives the chat replay with checkpoints and picks up again after an interruption
  • Local React player with the chat in sync with the video

Technologies: TypeScript, Node.js, React, hls.js, Tailwind CSS, Vitest, Playwright, GitHub Actions

Open on GitHub

See on npm

Antevue

Machine learning research project on football: it studies how to forecast matches better from professional event and tracking data, and runs every matchday unattended on Linux.

  • Professional data from commercial providers, the same kind clubs use: events, player tracking, lineups and xG
  • Idempotent data pipeline: every response is stored raw and reconciled until the provider stops correcting it
  • Per-season features validated point in time to avoid information leaks
  • Dixon-Coles and LightGBM models tuned with Optuna, with calibration and backtesting
  • Linux server with Docker and PostgreSQL, and every matchday automated with Airflow and systemd

Technologies: Python, LightGBM, Optuna, FastAPI, PostgreSQL, Apache Airflow, React, TypeScript, Docker, Grafana, pytest, GitHub Actions

The repository and the data are private. I can walk through the architecture or give a demo.

Education

Experience

Skills

Languages

Download the CV as PDF

Contact