Built a fault tolerant stream processing system with Redis backed checkpoint management, implementing automatic recovery mechanisms and state persistence for high availability data pipelines

Designed type-safe Scala APIs using advanced generics and Akka Streams, creating composable flow abstractions that ensure compile-time safety while maintaining runtime performance

Implemented comprehensive testing infrastructure with integration tests, mock checkpoint managers, and timing-sensitive test scenarios to validate distributed system behavior under various failure conditions

Created a Nix derivation for Playwright that allows for custom browser versions and extensions, enabling reproducible browser automation workflows

Implemented a flexible configuration system that supports multiple browser versions, custom extensions, and proxy settings, allowing for easy customization and extension of the Playwright environment

Built a comprehensive testing infrastructure with integration tests, mock browsers, and headless mode support, ensuring the reliability and stability of the Playwright environment on the server

Designed and implemented a full-stack quantitative trading simulator with a React frontend and a C++20 backend, enabling users to upload market data, configure trading strategies, and run backtests with real-time visualization of results

Built a responsive and interactive user interface using React, incorporating charting libraries for market data visualization and strategy performance analysis

Implemented a high-performance C++20 backend that handles large volumes of market data and complex trading strategies, ensuring fast and reliable backtesting results

Built a lightweight SSE client library for Cloudflare Workers and durable objects, enabling real-time event streaming from server to client in serverless environments

Implemented event buffering, reconnection logic, and custom event types to handle dynamic data streams with ease

Designed a modular architecture that allows for easy integration with existing Cloudflare Worker applications and services

Developed a comprehensive iOS productivity app using SwiftUI and Apple's Screen Time API - Built LockedIn, a focus enhancement application that integrates FamilyControls, ManagedSettings, and DeviceActivity frameworks to block distracting apps and track user behavior with detailed analytics including success rates, streaks, and screen time metrics

Implemented advanced timer systems and productivity features - Created a sophisticated Pomodoro timer with customizable focus/break cycles, manual lock timers, rule-based automation, and real-time progress tracking with force-quit detection and behavioral feedback systems

Built a complete widget ecosystem and monetization system - Developed multiple widget variants (small, medium, accessory) with live timer updates, integrated RevenueCat for subscription management with freemium/pro tiers, and implemented comprehensive data persistence using UserDefaults and App Groups for seamless widget-app communication

Developed a high-performance process manager in TypeScript/Bun that leverages Bun's native APIs for 2-3x faster process spawning compared to Node.js, featuring auto-restart policies, real-time monitoring, clustering support, and daemon-based architecture for production-grade process lifecycle management

Implemented advanced memory optimization techniques including circular buffers for metrics history, string pooling for duplicate data reduction, and custom memory tracking systems that prevent memory leaks and maintain constant memory usage regardless of runtime duration

Built comprehensive testing infrastructure with performance benchmarks, stress testing, integration tests, and memory optimization validation, achieving robust error handling with graceful shutdowns and automatic recovery mechanisms for enterprise-level reliability

Built a mobile app using React Native and Expo, enabling resellers to find the best deals on vinted by analyzing market data and providing real-time price comparisons and notifications

Implemented a robust backend using Cloudflare Workers and NginX on a VPS, handling user authentication, data storage, and API requests with D1, ensuring secure and scalable performance

Built a robust scraping service with Kotlin and a custom-built web crawler that efficiently scrapes vinted's market data, handling pagination, rate limiting, and data normalization

Designed a user-friendly interface with intuitive navigation and search filters, enhancing the user experience and driving app engagement