Builder from Vancouver πŸ‡¨πŸ‡¦

Software engineer with 7 years of experience and a master’s in CS. Born and raised in πŸ‡ΊπŸ‡¦, based in πŸ‡¨πŸ‡¦. Outside of work: snowboarding, concerts and travelling.

πŸ“¦ Projects

πŸ€–

Vibe Buddy

A desktop robot that shows usage limits for AI coding tools like Claude Code and Codex, and lets you know when they’re working or finished. I started it as a hobby project with Alize. It ranked #6 on Product Hunt’s daily leaderboard.

πŸ•ŠοΈ

Played in Russia

Public database tracking artists who perform in russia after the invasion of Ukraine. This is a personal project: I enjoy going to concerts and believe in sanctions and boycotts against countries that disregard the international order and invade others. My Reddit post about it reached 100,000 views, and the project was featured in Ukrainian and European media.

🍁

My Days in Canada

iOS app to track physical presence in Canada for citizenship eligibility, log trips, and optionally scan photos for travel dates β€” all data stays on-device/iCloud.

Show moreprojects Show lessprojects
πŸ™ˆ

The Shy Dock

A tiny macOS menu bar app that auto-hides your Dock when you're on the laptop alone and brings it back when an external monitor is connected.

πŸ₯”

Potato Classifier

ML-powered app that identifies potatoes. Because why not make something fun while practicing computer vision? I handled the engineering, while Alize worked on computer vision.

πŸ‘” Work

DeviantArt

2022 - 2026 Β· Vancouver, Canada

I developed APIs and data pipelines, ran A/B tests, trained and deployed machine learning models, built full-stack apps and worked with OLAP systems. My stack included AWS (CDK, Lambda, S3, Kinesis, SQS, SNS and Step Functions), Grafana, Kubernetes and Next.js.

I guided a junior developer on an A/B testing management project, designing the system’s architecture and supporting their implementation and QA work.

I independently led and built a high-load replacement for the legacy pipeline that processed every image upload and updates to files, descriptions and other content. I also developed image-processing APIs that ran ML models at 5–8 requests per second. I developed and led an internal marketing tool that sent millions of notifications to users each week.

SoftServe/Cisco

2021 - 2022 Β· Gdansk, Poland

Our team worked on email spam detection for Cisco Talos, Cisco’s threat intelligence team.

I built a cached translation API designed to handle large volumes of email, translating message content to help the system identify spam.

I wrote extensive technical documentation and worked with Python, Kubernetes, PostgreSQL, Redis and email protocols.

GoBoutique

2019 - 2021 Β· Lviv, Ukraine

I worked closely with traders to turn their market insights and hypotheses into algorithmic trading strategies. This involved collecting, processing and analyzing financial data, then translating trading ideas into code.

I also built chatbots and automated data pipelines using Kubeflow and AWS Step Functions to collect and prepare data for training machine learning models.

My stack included Python, Vue.js, AWS (S3, Lambda, EC2, Fargate, Step Functions), pandas, NumPy and Plotly.

πŸ“« Contact me

Check out my Twitter or GitHub profile

Or can reach me via [email protected]