Featured Projects · Production Work
Co-founded and scaled Poker Scientist from concept to 500 paying users, building a sophisticated platform that solved what the poker industry considered computationally impractical: making Nash equilibrium strategies understandable and actionable for human players.
This portfolio was built entirely using AI agents orchestrated through the BMAD-Method framework—not as a shortcut, but as a deliberate exploration of emerging development practices. As a progressive developer, I recognized agentic development as a highly relevant technology evolution. Rather than theorizing, I chose to learn by doing: building a production-grade portfolio while mastering agent-driven workflows.
AI Companion is a conversational AI assistant designed to provide personalized, context-aware interactions through persistent memory and intelligent knowledge retrieval. Built with React and OpenAI GPT-4, the platform implements Retrieval-Augmented Generation (RAG) architecture to maintain long-term conversation context, user preferences, and domain-specific knowledge across sessions. Unlike stateless chatbots, AI Companion remembers past interactions, learns user communication patterns, and surfaces relevant historical context to deliver increasingly personalized and valuable assistance over time. The system integrates vector databases (Pinecone) for semantic memory storage, Redis for session management, and WebSockets for real-time bidirectional communication, demonstrating practical AI/ML implementation beyond basic API integration.
