High-performance live algorithmic trading platform capable of processing 50,000 ticks/second. Multithreaded and event driven with backtesting and optimisation tools for a comprehensive strategy development environment. Achieved 9% returns with <2% drawdown with deployed strategies so far in Q4 2024.

High-performance market data streaming infrastructure built with Spring Boot, Project Reactor and Kafka. It cleanly integrates with new brokers with initial support for Oanda. Features include: real-time price streaming with Protobuf serialization, resilient reconnection handling, micrometer metrics, and a REST API for stream management. Part of a larger trading ecosystem, designed for multi-broker support and high throughput.

A markdown-to-Lua transpiler enabling literate programming for Neovim configurations. Features LSP integration for real-time error reporting, custom pragma directives, and precise line number preservation. Built with Go using the goldmark parser, allowing developers to maintain readable, documented lua configurations while preserving full IDE support.

An Ngrok alternative for exposing local ports. Automatic service discovery based on docker-compose, allowing spinning up full applications and exposing them temporarily to the internet, for testing, debugging and demoing purposes.

Automated Python application transforming Reddit posts into TikTok videos. Dynamic subtitle creation, audio synthesis, and Celery/Redis task queuing allows for automated viral content gaining hundreds of views (see tiktok.com/@ReddReelStoryz.)