Michael Fang
Co-Founder at Spectacular Labs
Berkeley, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMichael Fang is a PhD-trained founder and engineer in Berkeley with 11 years of experience applying physics, mathematics, statistics and AI to real-world problems—from optical and photonic neural hardware to autonomous scientific instrumentation and food safety. As Co-Founder of Spectacular Labs and a former UC Berkeley researcher, he combines deep-domain research in analog/photonic acceleration and surrogate modeling with hands-on systems engineering and startup execution. His open-source work includes security hardening for Jupyter Notebook and backend contributions to Letta (formerly MemGPT), adding debugging, FAISS indexing and CSV/PDF preloading for stateful LLM agents. Comfortable bridging lab prototypes and production services, he brings a rare mix of experimental physics insight and pragmatic software engineering to accelerate ML-hardware and automated science applications.
11 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD, Physics, Doctor of Philosophy - PhD, Physics at University of California, BerkeleyBachelor of Science - BS, Physics & Mathematics, Bachelor of Science - BS, Physics & Mathematics at University of Toronto
English, Chinese, French
Github Skills (18)
debugging10
debug10
python10
jupyter10
it-security10
security10
jupyter-notebook10
system-security10
faiss9
api8
agent8
ai8
llm8
apim8
api-design8
Programming languages (10)
Github contributions (5)
letta-ai/letta
Oct 2023 - Nov 2023
Letta (formerly MemGPT) is the stateful agents framework with memory, reasoning, and context management.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:4 releases, 68 reviews, 83 PRs in 28 days
Contributions summary:Michael primarily focused on debugging and extending the functionality of the `letta-ai/letta` repository, which is a framework for stateful agents. Their commits introduced a debug flag and debugging output to the `interface.py`, `main.py` and `memgpt/utils.py` files, facilitating more detailed monitoring. Moreover, the user implemented necessary scripts to generate a FAISS index, and added functionality for supporting preloading files. Furthermore, they incorporated CSV and PDF support for preloading.
aiai-agentsllmllm-agent
jupyter/notebook
Feb 2016 - Apr 2016
Jupyter Interactive Notebook
Role in this project:Back-end Developer
Back-end DeveloperContributions:6 commits, 2 PRs, 3 comments in 1 month
Contributions summary:Michael primarily focused on security-related improvements within the Jupyter Notebook application. Their contributions involved preventing the notebook from running as a root user by implementing checks for user privileges across different operating systems. They added and modified code to ensure proper handling of root user access, addressing potential security vulnerabilities and improving the application's robustness. Additionally, the user introduced a feature to toggle line numbers within the notebook interface.
jupyter-notebooknotebookjupyterlabnotebooksjupyter
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.