👋 Hi, I'm Kenneth Wang

🚀 About Me

Principal-level Machine Learning Engineer working on AI agents, agent memory and reasoning, and large-scale pricing & marketplace ML systems.

🔗 Connect With Me

📱 Social Media

• LinkedIn: Kenneth Wang
• Twitter/X: @BichengW81192
• GitHub: @bichengwang
• Medium: Kenneth Wang

🎯 Professional Networks

• ResearchGate: Kenneth Wang
• Google Scholar: Kenneth Wang

🎨 Creative & Content

• YouTube: Kenneth Wang
• Twitch: kenneth
• Instagram: @kenneth
• TikTok: @kenneth

📧 Contact

• Email: bichengwang17@gmail.com
• Website: bichengwang.github.io

🛠️ Technologies & Skills

• Languages: Python, Go, Java, C++
• LLM / Agents: vLLM, LangChain, LlamaIndex, AutoGen, DeepEval, agent orchestration, agent memory, agent evaluation
• ML Frameworks: PyTorch, Horovod, Ray
• Data / Streaming: Spark, Flink, Kafka
• Infra: Kubernetes, Yarn, Redis, Cassandra
• Domains: Pricing, promotions, marketplace optimization, ranking, causal modeling, experimentation, simulation

📊 GitHub Stats

🌟 Patents & Research

Patents (ML)

• CN105068743B
• CN104966016B
• CN104850773B

Research

• Stanford — Reinforcement Learning on Stock Trading: Portfolio strategies combining auto-regressive forecasting with RL (PPO, A2C, DDPG, TD3, SAC). Beat the S&P 500 in rolling backtests with 1.33 Sharpe, 15% alpha, beta = 0.9.
• UC Irvine — Recommendation System & Data Storage: Recurrent PageRank on Hadoop MapReduce using adjacency-matrix and embedding representations for realtime user and content recommendation.

📈 Current Focus

• 🤖 Agent memory, reasoning, and personalization for enterprise AI agents
• 📊 Large-scale pricing, promotion, and marketplace ML systems
• 🔬 LLM fine-tuning (SFT / RL), agent evaluation, and data flywheels
• 📚 Continuous learning and open-source contributions

---

Feel free to reach out if you'd like to collaborate on interesting projects! 🤝