Ivan Mustière
Live

Interactive projects

No screenshots, no videos: launch a project, a real Docker container starts on my server and you get the interface within seconds, nothing to install.

  • 256 MB – 1.5 GB of RAM
  • Half a CPU core
  • No internet access
  • Destroyed after 15 min
  • Pinned image versions

Available demos

How it works

  1. 01

    Pick a demo

    Each project is a real service, not a screenshot: you use it just like in production.

  2. 02

    A container starts

    The backend runs a dedicated docker compose on my VPS. A second visitor joins the same instance.

  3. 03

    Explore for 15 min

    The interface opens in the page. You can extend 3 times, then everything is destroyed automatically.

Want to see the architecture and the live server status? Under the hood →

Other projects

Built at work, at school or on my own, without an interactive demo. Click a technology to filter.

9 projects shown

PRTG & Grafana alerting automation

Work

Custom PRTG sensors and Grafana dashboards to automatically detect and report major infrastructure issues.

  • Custom PRTG sensors
  • Grafana dashboards
  • Automated alerting

Data batch management

Work

Orchestrating and monitoring batch jobs with GoAnywhere: automated data flows with error handling and restarts.

  • Batch scheduling
  • Run monitoring
  • Error handling and restarts

PostgreSQL → Snowflake migration ETL

Work

End-to-end ETL pipeline migrating and managing a PostgreSQL warehouse’s data in Snowflake, with transformation and validation.

  • Full schema migration
  • Transformation and validation
  • Reproducible pipeline

Public transport feed aggregator

School

Web platform gathering real-time public transport data from across France: RATP (Paris), NaoLib (Nantes), TCL (Lyon) and other networks.

  • Multi-network integration
  • Real-time data
  • Unified map interface

Paris road traffic prediction

School

Software predicting traffic on every road in Paris based on events, time of day and historical data.

  • Per-segment prediction
  • Event-aware
  • Trained on historical data

Badminton competition statistics

School

Statistical analysis of competition performance, identifying each player’s strengths and weaknesses across their rallies.

  • Strength/weakness profile per player
  • Interactive visualisations
  • Rally sequence analysis

Esports performance statistics

Personal

Statistical analysis of a video game’s characters to rate their strength in the current meta and build a data-driven tier list.

  • Data-driven tier list
  • Cross-patch comparison
  • Performance dashboard

CNN — Minimap from a video feed

Personal

Real-time generation of a game minimap by analysing the video feed with a convolutional neural network.

  • Frame-by-frame analysis
  • Real-time map reconstruction
  • Trained on game captures

Crowd movement flow analysis

Personal

Crowd movement prediction model inspired by Fouloscopie, analysing and anticipating pedestrian flow dynamics from videos.

  • Pedestrian flow detection
  • Movement prediction
  • Density visualisation

Got a project in mind?

I’m open to permanent or freelance opportunities.

Get in touch