Jiayi Wei

Jiayi Wei 家一 魏

Research Scientist at Augment

Augment Computing, Inc.

Biography

I’m a founding member and research scientist at Augment Computing, Inc., where I build innovative AI-powered developer tools. I earned my PhD in Computer Science from UT Austin under the guidance of Işıl Dillig and Greg Durrett, focusing on blending deep learning with static analysis to automate programming tasks like type inference and code autoediting. Along the way, I also gained experience in software security, verification, program synthesis, and robotics. Before UT, I earned my bachelor’s in Physics from the University of Science and Technology of China.

At Augment, I’ve worked on projects ranging from code completion to retrieval context, and I’m currently leading our research on the Next Edit feature. I’m passionate about building tools that empower developers, and I thrive on tackling complex AI challenges that push the boundaries of what’s possible in software engineering with large language models.

Interests
  • AI for Code
  • Software Engineering
  • Programming Language Theories
Education
  • PhD in Computer Science, 2018—2023

    University of Texas at Austin

  • BSc in Physics, 2013—2017

    University of Science and Technology of China

Recent Publications

(2023). Coeditor: Leveraging Repo-level Diffs for Code Auto-editing. In ICLR 2024 (Spotlight paper).

PDF Cite Code Video

(2022). TypeT5: Seq2seq Type Inference using Static Analysis. In ICLR 2023.

PDF Cite Code Video

(2022). STEADY: Simultaneous State Estimation and Dynamics Learning from Indirect Observations. In IROS 2022.

PDF Cite Code Video

(2021). OneVision: Centralized to Distributed Controller Synthesis with Delay Compensation. In IROS 2021.

PDF Cite Code Video

(2020). LambdaNet: Probabilistic Type Inference using Graph Neural Networks. In ICLR 2020.

PDF Cite Code Video