Jacob Berger

AI Researcher & Software Engineer
jacob@bergerj.com Full Resume GitHub
Experience
Lexel Synergetics
Portland, OR
  • Architected end-to-end data pipelines for training specialized LLMs.
  • Led and co-authored the approved Phase I SBIR proposal.
  • Deployed cloud infrastructure on AWS, GCP, and Vercel.
View Experience Details →
Research
Dynamic Constraints for LLMs
Honors Thesis
  • First-author submission under review at AAAI 2027 ("Parsimonious Decoding").
  • Developing novel metrics to enforce rigorous semantic constraints in probabilistic language models to improve task reliability.
  • Advisor: Dr. Lizhong Chen, OSU STAR Lab.
View Research Details →
Project
RelCube - Deep RL Rubik's Solver
Transformer Architecture
  • Transformer-based solver using structured piece representation.
  • Exploits physical cube geometry for SOTA efficiency (~3.5M params).
  • Trained with Approximate Value Iteration and Expanding Curriculum.
  • Team of 3 final project
View Project Details →
Education
PhD in Artificial Intelligence - Oregon State University
Exp. 2026-2030
Advisor: Dr. Lizhong Chen, STAR Lab
Honors B.S. in Applied CS & Math - Oregon State University
2022-2026
Summa Cum Laude • Honors Degree • AI Concentration • Double Major • GPA: 3.9+
Core Skills
Python C/C++ PyTorch LLM Finetuning LLM Serving Linux/Debian Virtualization (Proxmox/KVM) AWS/GCP Full Stack Web
Browse All Projects