Summary
Zhiyuan Guo is a research-oriented software engineer and CMU student with eight years of hands-on experience building full-stack systems and ML research infrastructure. At CMU’s Machine Learning Department he developed Adaptive Asking, an offline RL framework that lets LLM planning agents decide when to query for missing information, and built the PyTorch/HuggingFace evaluation pipelines that benchmarked GPT-4, GPT-3.5, and Llama-2-70B. He also contributes to education-focused open-source engineering, having been a major developer on P5play V3—a JavaScript physics/graphics engine used by CodeHS, Code.org and universities. Comfortable across front-end mocks to backend experimentation, he blends product-minded prototyping (Uber case winner) with rigorous research execution. Based in Pittsburgh, he’s driven by practical impact—“don’t stop”—and a knack for turning ambiguous problems into measured, reproducible gains.
8 years of coding experience
1 year of employment as a software developer
Middle School Diploma, Middle School Diploma at Massachusetts Institute of Technology
Harvard University
High School Diploma, High School Diploma at Southridge School
Bachelor's degree Statistics + Machine Learning, Bachelor's degree Statistics + Machine Learning at Carnegie Mellon University
English, Chinese, Spanish, French