I build high-performance trading engines and scalable computer vision pipelines. Focused on C++20, Python, and Cloud Infrastructure.
The Challenge: Visualizing complex market dynamics and testing strategies in a responsive, real-time environment using synthetic data.
The Solution: Built a comprehensive trading simulation from scratch using C++20 and Dear ImGui. Features procedurally generated market data and a modular CMake-based architecture.
void TradingApp::RenderUI() {
ImGui::Begin("Market Dashboard");
if (ImGui::Button("Generate New Tick")) {
engine.simulate_tick();
}
// Custom ImGui plotting for price history
DrawPriceChart(engine.get_history());
ImGui::End();
}The Challenge: Achieving precise physical dimension estimation from 2D images by using known reference objects for scale.
The Solution: Developed a computer vision pipeline for image-based size estimation via reference objects. The entire application is hosted on AWS for performance and scalability.
The Challenge: Managing and switching between various local and cloud-based Large Language Models (LLMs) from a single, intuitive interface.
The Solution: Developed a TypeScript-based UI chat interface supporting local LLMs via Ollama and integrating cloud LLMs (e.g., OpenAI, Anthropic) via API keys. Features dynamic model switching and real-time response streaming.
Currently open for internships and junior positions.