Full-stack Developer in Vancouver, Canada

An online tool to quickly generate beautiful and unique word art designs. Users can choose shape, words, fonts, colors, and layout and the app will fill the shape to create an aesthetically pleasing design. The results can be exported as high-quality images or as SVG for later printing for posters, canvases, or merchandise.

It's my hobby indie project and I'm the creator and the only developer.

Technologies: Rust, Web Assembly, TypeScript, React, NestJS, TypeORM, PostgreSQL

This is a personal project of mine that involves an online microphone tuner for guitar and other instruments. It's likely the fastest and most accurate online microphone tuner out there. It is internationalized in English and Russian.

Technologies: Web Audio, Gatsby, React, TypeScript, Rust, Web Assembly, HTML Canvas

The app helps music students (mainly jazz students) to practice better by automatically generating challenging practice drills. It also can help music teachers to create drills for their students.

Users can create sharable practice sessions, tweak session parameters to fit their needs (e.g., by adding chords, scales, random note sequences, etc.), play the exercises back, export them as MIDI, and more.

Technologies: Firebase, React, MobX, Web Audio API, VexFlow

About the Reporter Suite app:
This is the app of choice for US court reporters and it greatly simplifies their day-to-day routines.

Some Features:
• Invoice management system with built-in payment. processing time, expenses, case log tracking, and more
• Generating official reports based on the tracked data.
• Public shareable profile pages for court reporters with contact and order forms.
• Social features.

My Role:
My client asked me to do a complete rewrite of the AngularJS legacy app running on the Stamplay platform to something more modern, maintainable, and scalable and also to expand the app functionality based on the feedback from users of the old version.

Work Highlights:
• Set up the AWS infrastructure, the staging environment, monitoring, user analytics dashboards, and error tracking; conducted load-testing.
• Added Stripe payments processing.
• Enabled authentication with Google accounts
• Added new user roles and user interactions: e.g., users receive notifications of incoming invoices that other users create for them, etc.
• Added an admin UI to simplify app maintenance.

Technologies: AWS, Docker, React, Redux, PostgreSQL, Node.js, Hapi, Sequelize, Metabase, TrackJS, Stripe, Redis

The new mobile-first real-time web app for Smarkets, a London-based prediction market (aka betting) exchange.
Being a part of a very tight team with daily standup meetings and frequent cross-team communication, there was a lot of responsibility on my shoulders and I contributed a great deal to the app's functionality and performance.

The project itself is meant to replace the older monolithic legacy website. It now has a much more maintainable code base (some parts of which are shared with our React Native mobile app), a top-notch technology stack, and is built upon the best practices for modern high-performant mobile sites.

This project is an online contest platform for high-frequency trading strategies, simulated on real historical data. The idea is to popularize (especially among students) the traditionally secretive field of HFT (high-frequency trading) to spot and attract talent to join the company's quantitative analyst pool.

Participants can implement their trading strategies in either C++ or Python, and then run it either locally with a provided SDK (for debugging) or use the web-interface to schedule their simulations to be run on a cluster infrastructure. After the simulation is finished, they can see charts visualizing the trading behavior of their algorithms, test and compare different approaches, and export the data for further analysis.

I started working on this project on my own (including the initial prototype and the refined microservice architecture), and after several months, two people joined in and we've been working together on it since then. I designed the back-end architecture (a pool of simulation nodes, microservices, a storage layer, and message queues between them), developed the API and the web-interface for participants (Django/Flask, JavaScript), and took care of the infrastructure, CI, and automation.

A demo SPA (front-end only) which is a drag-n-drop visual editor for back-end logic, similar to micro:bit from Microsoft. Users could use visual blocks, rearrange (with possible nesting), and configure them so that they express required business logic.

I've implemented only the front-end part of the project with the back-end being mocked for the demo purposes.

Technologies: React, Redux, Vanilla JavaScript, ES6, Webpack, AWS, Docker

A library that implements CART decision trees as well as lesser-known Oblivious Decision Trees for classification and regression problems; it's fully compatible with the popular ML package Scikit-learn.

This project is the result of my interest in machine learning—implementing decision trees from scratch allowed me to deeply understand the concept.

A two-week test project done as a part of the Toptal onboarding process.

This is a multi-user time tracker SPA. I've implemented both the front- and back-end following the project requirements and working on it during the evenings and the weekends.

Test credentials:
login: user@test.com
password: password

Technologies: React, Redux, MongoDB, Sequelize, Heroku

A web app and a chatbot for tracking income/expenses, suitable for personal use or for collaboration in small groups (e.g., families, or a joint holiday trip with friends). It creates the visualization of your spending habits and allows the importation and exportation of data for further analysis.

Technologies: Node.js, LoopBack.io, ES6, Webpack, React, Redux, HTML5, Sass, Python 3, PyMongo, MongoDB, AWS, Docker, Linux, Bash Script

A post explaining the approach that I took while working for Smarkets in supporting the different color themes in their web application.