🎓 Research-oriented Developer
🤖 Reinforcement Learning · LLM · AI Agents · RAG
🏥 Intelligent Medical & Decision-Making Systems

I focus on reproducible AI research and system-level implementations, bridging the gap between academic ideas and practical, deployable systems.

• Reinforcement Learning & Sequential Decision Making
• LLM-based Agents and Tool-using Systems
• Retrieval-Augmented Generation (RAG)
• AI-assisted Medical Question Answering
• Human-AI Interaction, Embodied Intelligence, and System Design
• Super-individual AI systems: amplification of human capability through AI agents

I am particularly interested in how embodied intelligence and agent-based systems can enable super-individual capabilities—AI systems that do not replace individuals, but significantly amplify human perception, decision-making, and long-term agency through interaction with environments, tools, and other agents.

• 🧪 Reproducing and analyzing RL papers with controlled experiments
• 🤖 LLM-driven decision agents with memory and tool use
• 🏥 AI-powered medical consultation systems (RAG + structured reasoning)
• 🌐 Full-stack systems for AI-assisted collaboration and mentoring

Languages
Python · JavaScript · TypeScript

AI / ML
PyTorch · NumPy · RL Algorithms · LLM APIs

Backend / Systems
Node.js · Docker · REST APIs

Frontend
React · CSS · UI Prototyping

Tools
Git · Linux · VS Code · LaTeX

• 🔬 RL Reproduction Projects – faithful re-implementations and ablation studies
• 🤖 LLM Agent Systems – planning, memory, and tool-augmented agents
• 🏥 Medical AI Systems – RAG-based intelligent consultation demos
• 🌐 Full-stack Projects – React + Node.js applications

(See pinned repositories below 👇)

• Reproducibility over hype
• Systems thinking over isolated models
• Long-term impact over short-term metrics

• GitHub: @MurphySong