Shangyin Tan
Research Intern at Databricks
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolShangyin Tan is a research-focused software engineer and PhD candidate at UC Berkeley specializing in building compound AI systems and stateful agents, currently interning at Databricks after research roles at Letta, DeepMind, and Microsoft Research Asia. With eight years of experience spanning program analysis, functional programming, and ML-driven config synthesis, he bridges programming language theory and practical agent design. His contributions to DSPy emphasize programmatic assertions and robustness for chain-of-thought and retrieval-augmented workflows, reflecting a preference for programming models over prompting. Based in Palo Alto, he combines academic rigor with applied research in memoryful agents and agentic evaluation. He has a track record mentoring undergraduates in competitive programming and systems courses, signaling strong teaching and engineering communication skills. Notably, his work often focuses on making AI systems auditable and reliable through formal validation mechanisms rather than opaque heuristics.
8 years of coding experience
1 year of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at University of California, BerkeleyNanjing Foreign Language SchoolBachelor of Science - BS Computer Science Honors, Bachelor of Science - BS Computer Science Honors at Purdue University
Stackoverflow
1reputation
0reached
1answer
0questions
Github Skills (7)
machine-learning10
assertion-framework10
dsym10
python10
retrieval-augmented-generation9
testing8
google-colaboratory6
Programming languages (9)
Github contributions (5)
stanfordnlp/dspy
Sep 2023 - Jan 2025
DSPy: The framework for programming—not prompting—language models
Role in this project:
ML Engineer
ML EngineerContributions:3 reviews, 18 PRs, 52 pushes in 1 year 4 months
Contributions summary:Shangyin's commits focus on adding and modifying assertions within the DSPy framework. These changes involve implementing assertion mechanisms and incorporating them into the core program logic, specifically related to chain-of-thought reasoning and retrieval-augmented generation. The modifications suggest an effort to improve the robustness and reliability of DSPy programs. The code additions provide functionality for validating search queries and incorporating feedback mechanisms within the program's architecture.
nlpbertknowledgepredictlanguage-models
Shangyint/Shangyin-Tan-Resume
Nov 2021 - Jan 2025
Contributions:9 pushes, 3 branches in 3 years 2 months
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.