Junlajak Jongpipattanakul (Nun)
Tutoring
I tutor physics and math.
I just finished my MS EECS at UC Berkeley. I'm an ex-IPhO Thailand rep tutoring physics and math part-time. Online or onsite.
01
Education
MS, Electrical Engineering & Computer Science
May 2026 GPA 4.00 / 4.00 · ML hardware · AI infrastructure Thesis Thesis · HZO FeFET neural networks
BA, Physics & Computer Science (double major)
May 2025 GPA 3.94 / 4.00
02
Test scores
US standardized
- SAT Math
- 800 / 800
- SAT Subject — Math II
- 800 / 800
- SAT Subject — Physics
- 800 / 800
- AP Calculus BC
- 5 / 5
- AP Physics C: Mechanics
- 5 / 5
- AP Physics C: E&M
- 5 / 5
- GRE Quantitative
- 170 / 170
Thai admissions
- PAT1 (Mathematics)
- 288 / 300
- 9 Wichasaman Physics
- 100 / 100
03
Olympiad
Physics
- IPhO
- Thailand Rep.
- APhO
- Thailand Rep.
- EuPhO
- Bronze
- TPhO
- Gold
Astronomy
- IAO
- Silver
- TAO
- Gold
04
Experience
Student Learning Center
Physics Tutor · Jan 2023 – Jun 2024 · Berkeley
- Three semesters of drop-in tutoring, ~10 hours per week — Physics 5A (mechanics) and Physics 7B (E&M, thermodynamics, waves).
- Supported roughly 20 students per week.
PhysElo
physelo.com- Built a competitive physics tournament platform end-to-end (Next.js, Supabase, TypeScript) with live contests, real-time leaderboards, and free-response grading, architected to support 100+ concurrent users.
- Designed an algorithm-driven scoring engine with continuous time-decay and attempt penalties, and engineered a Codeforces-style Elo ranking system featuring zero-sum point exchange math.
- Built a multi-agent LLM pipeline that drafts olympiad-level physics problems against a benchmark corpus, generates TikZ diagrams, and self-verifies solutions.
Studata Solution Co.,Ltd
Full-Stack Developer (Part-Time) · June 2022 – Aug 2024 · Remote
- Architected a learning management system web application for SAT prep with AI-driven recommendations and a feedback system using React/Next.js and Django REST Framework.
- Implemented graph algorithms for creating personalized exams based on users' past performances and integrated OpenAI APIs to help generate exams based on model exams, cutting instructors' workload by 50%.