Jefferson Lin

I am pursuing a B.S. in Artificial Intelligence at MIT, focusing on reinforcement learning, robotics, and agent systems. I currently work at CSAIL and Harvard’s AI & Robotics Lab on improving hierarchical agentic goal decomposition, imitation learning, and OpenVLA techniques. To continue my studies, I turned down my acceptance to Y Combinator’s Fall 2025 batch by David Lieb for an agentic ML project I worked on in high school.

Before MIT, I contributed to CNNs at Dr. Dena J. Clink’s lab at Cornell and helped improve OpenFace’s accuracy at Carnegie Mellon. I also developed early AI voice-agent systems, end-to-end neural network applications, full-stack platforms, and MLOps automation pipelines on my own, while gaining industry experience as a SWE intern at LexisNexis Risk Solutions through CodeDay Labs.

Some noteable things I did:

  • Turned down YC acceptance after graduating high school
  • Ranked top 5 in North America for PUBG mobile in middle school
  • Built Dormsell, MIT's first web marketplace
  • Started my high school's first ever engineering club and chemistry olympiad team
  • Founded Simply Code, a 501(c)(3) nonprofit bringing hands-on technology education to 100+ students
  • Wrote a first edition 300+ page AI textbook using AI in three hours to demonstrate AI's potential
  • Met with Rep. Jared G. Solomon to talk about local AI policy

Motivated by my upbringing in a low-income immigrant family, I’m driven to build AI and robotics systems that are safe and widely accessible. Feel free to reach out, I’d love to learn more about your work and potentially collaborate!


Relevant Experiences
Researcher
Researcher

Jefferson Lin, Henry Lieberman, David Canfield Smith, Hugo Sanchez, Honjar Xing

MIT Computer Science & Artificial Intelligence Laboratory 2025 - Present

Researching new agentic hierarchical imitation learning and goal decomposition techniques.

Researcher

Jefferson Lin, Henry Lieberman, David Canfield Smith, Hugo Sanchez, Honjar Xing

MIT Computer Science & Artificial Intelligence Laboratory 2025 - Present

Researching new agentic hierarchical imitation learning and goal decomposition techniques.

Machine Learning Researcher
Machine Learning Researcher

Jefferson Lin, Librecorder Group

Public Invention 2025

Contributed to a low-cost, open source microscope diagnostic system to enable rapid disease detection in rural areas.

Machine Learning Researcher

Jefferson Lin, Librecorder Group

Public Invention 2025

Contributed to a low-cost, open source microscope diagnostic system to enable rapid disease detection in rural areas.

Full Stack Developer in Startups
Full Stack Developer in Startups

Jefferson Lin

Y Combinator Fall Batch 2025

Designed an agentic MLOps diagnostic tool, being invited to AI Startup School with portfolio of full stack apps.

Full Stack Developer in Startups

Jefferson Lin

Y Combinator Fall Batch 2025

Designed an agentic MLOps diagnostic tool, being invited to AI Startup School with portfolio of full stack apps.

Software Engineering Intern
Software Engineering Intern

Jefferson Lin, David Zapata

LexisNexis Risk Solutions 2025

Reduced codebase inefficiencies with SLOCCount integration in HPCC Systems using the ECL programming language.

Software Engineering Intern

Jefferson Lin, David Zapata

LexisNexis Risk Solutions 2025

Reduced codebase inefficiencies with SLOCCount integration in HPCC Systems using the ECL programming language.

Assistant Researcher
Assistant Researcher

Jefferson Lin, Yang Center SE Asia Research

K. Lisa Yang Center 2025

Researched CNN with geographic priors for Gibbon call classification and annotated 1000+ bioacoustic spectrograms under Dr. Dena J. Clink's Lab.

Assistant Researcher

Jefferson Lin, Yang Center SE Asia Research

K. Lisa Yang Center 2025

Researched CNN with geographic priors for Gibbon call classification and annotated 1000+ bioacoustic spectrograms under Dr. Dena J. Clink's Lab.

Student Researcher
Student Researcher

Jefferson Lin, Team Research Lab at Carnegie Mellon University

Carnegie Mellon University, Governor's School 2024

Led the mmWave sensor team in a multi-modal system that improved Open Face's facial detection software from 70.3% to 99.2%.

Student Researcher

Jefferson Lin, Team Research Lab at Carnegie Mellon University

Carnegie Mellon University, Governor's School 2024

Led the mmWave sensor team in a multi-modal system that improved Open Face's facial detection software from 70.3% to 99.2%.

Product Manager
Product Manager

Jefferson Lin, International Team

Cognitive Classrooms Pitch (NEOM & NY Junior Sciences Academy) 2023-2024 1st Place Winner

Developed an AI app for personalized learning with unique learning styles using LLMs, ElevenLabs, and MidJourney, placing 1st out of 142 teams.

Product Manager

Jefferson Lin, International Team

Cognitive Classrooms Pitch (NEOM & NY Junior Sciences Academy) 2023-2024 1st Place Winner

Developed an AI app for personalized learning with unique learning styles using LLMs, ElevenLabs, and MidJourney, placing 1st out of 142 teams.