Serge Serbinenko

Vancouver, BC ·

Machine learning engineer in Vancouver. I build AI tools and native apps, and I like software that is calm, fast, and good to look at.

See the workRésuméSay hello

01 — Now

Right now

By day I am a machine learning engineer at PressReader, the platform behind thousands of newspapers and magazines worldwide. I also build a broadcast-monitoring platform for a private client.

The rest of the time I build the tools I want to use, mostly in TypeScript, Swift, and Rust, with coding agents doing a lot of the typing. The places below are a tour of that.

  • Based inVancouver, BC
  • Working atPressReader
  • SpeaksEnglish, Russian, French
  • Off the clockCoaching track & field

02 — Work

Places I’ve built

Every project here gets its own place. Keep scrolling to travel between them.

All projects
T3 Pretty showing a new thread over a blurred photo of Los Roques, Venezuela

01 / 05 · Fork · 2026

Status: In daily use

T3 Pretty

The coding-agent app I work in every day: a fork of T3 Code with a real place behind every thread.

  • TypeScript
  • React
  • Electron
  • Swift
Read the storyGitHub
Spotiurge playing an album, its track list floating over a dimmed mountain-lake photo

02 / 05 · Fork · 2026

Status: In development

Spotiurge

A fast, native Spotify client in Rust, extended with discovery, sync, and album-matched scenery.

  • Rust
  • egui
  • librespot
  • Swift
Read the storyGitHub
Wraplab studio showing a car in a purple colour-shift wrap next to a panel of film swatches

03 / 05 · Product · 2026

Status: In development

Wraplab

See a real wrap film on your car in 3D before you pay for it.

  • TypeScript
  • React
  • Three.js
  • Node.js
Read the story

04 / 05 · Product · 2026

Status: Private beta

Storylight

Spoiler-safe scene art for audiobooks, revealed only when you reach the scene.

  • Node.js
  • Chrome extension
  • EPUB
  • Image models
Read the story

05 / 05 · Client work · 2025–2026

Status: In production

Broadcast monitoring

Round-the-clock radio and TV monitoring: live transcripts, AI-detected segments, and one dashboard to search them.

  • Python
  • TypeScript
  • Speech-to-text
  • LLMs
  • Google Cloud Run
Read the story

03 — Résumé

The route so far

  1. 2024 — Now

    Machine Learning Engineer

    PressReader

    Machine learning for a platform that delivers thousands of newspapers and magazines worldwide.

  2. 2025 — Now

    Lead Engineer (Contract)

    Broadcast monitoring platform, private client

    Round-the-clock radio and TV monitoring with live transcripts, AI segment detection, and one dashboard.

  3. 2016 — 2019, 2023 — Now

    Software Developer

    Sapphire4U

    Family software studio working on freelance projects.

  4. 2023 — 2023

    Software Engineer (Contract)

    EveryState

    Built an AI real-estate assistant on GPT-4 function calling.

  5. 2020 — 2021

    System Administrator

    PressReader

    Network Operations Centre for a worldwide newspaper and magazine platform.

Full résumé

04 — Contact

Send a postcard.

Tell me what you are building, or just say hello.

Write to me