Full-stack developer (Nest.js, Next.js) - Pulsecheck
(2026-07 - 2026-08)
A microservices system for monitoring endpoint uptime and performance, with real-time alerts.
- Designed and built a 4-service architecture communicating via RabbitMQ, with NestJS on the backend and Next.js on the frontend
- Implemented a scheduled ping worker with Redis-based distributed locking to prevent duplicate checks across instances, writing time-series results to MongoDB
- Built a real-time dashboard using WebSocket for live monitor status and response-time activity charts
- Added multi-channel notifications (email via Resend, Telegram bot) triggered on monitor status change through a RabbitMQ event pipeline
- Wrote unit and e2e test coverage (Jest, Supertest) across all backend services
- Set up CI/CD with GitHub Actions (automated testing and builds on every PR, branch protection requiring passing checks) and containerized all services with Docker; deployed to Railway (backend) and Vercel (frontend)
Full-stack developer (React Native, TypeScript, Python/FastAPI) - Wroom-Wroom
(2026-03 - 2026-06)
Smart mobile assistant for real-time vehicle diagnostics and automated maintenance tracking.
- Developed a cross-platform mobile application from scratch using React Native, Expo, and TypeScript, with real-time vehicle anomaly detection processing OBD-II sensor data (speed, RPM, battery voltage) every 200ms via BLE integration
- Implemented secure client-server communication with Axios auto-refresh token interceptors and encrypted local storage via Expo SecureStore
- Built backend functionality on Python/FastAPI with SQLAlchemy, including database schema modifications and new endpoints for a car inventory module (documents, fire extinguisher, first aid kit — with location and expiration tracking) and a service-station (СТО) visit log/diary
- Simplified navigation of complex, dynamic user flows with app-wide state management via Zustand and Expo Router.
Full-stack developer (Python/FastAPI, Next.js) - CalioT
(2024-11 - 2026-01)
AI system for monitoring the maturity of champignons on mushroom farms.
- Modernized the project's server infrastructure by migrating from Python/Django to Node.js/Nest.js, connecting microcontrollers, an AI-based computer vision pipeline, and a web application.
- Developed and implemented a PostgreSQL database schema for efficient storage and retrieval of camera data, images, and analysis results.
- Configured and integrated cloud storage using Azure Blob Storage, separating raw and AI-processed images to support scalability and performance.
- Enabled automatic maturity detection using a computer vision model in the analysis pipeline, achieved ~90% on training datasets.
- Enabled farmers to see crop analysis via a React frontend that displays camera data processed by a specially trained AI module.
- Developed firmware to enable communication between the ESP32-CAM microcontroller and the server, including the function of automatic image capture and photo upload.