Back to Template Library
Application Dossier · 2025

Michael
Nam

NationalityVietnamese
Age18 years old
SchoolHanoi-Amsterdam High School
ProgramBachelor of Science — Data Science & AI
Contactmichael.nam@student.edu.vn
GitHubgithub.com/michaelnam-ai
Country
France

Selective stretch lane rather than the affordability base.

Univ
École Polytechnique

Legacy sample packet for a top-end France target.

Cost
Scholarship-first

Only rational with major funding or a confirmed family budget.

Certainty of applying
Selective

Strong upside, but not one of the repo's default anchor applications.

3.98
GPA / 4.0
Top 1%
National Rank
Gold
Olympiad · Math
12+
GitHub Projects
DOC-ID: MN-POLY-2025-008 · CONFIDENTIAL
Target Institution · France
École Polytechnique
Bachelor of Science — Data Science & AI
PALAISEAU, FRANCE · EST. 1794
01 —

Motivation Letter

15 September 2025

Admissions Office
École Polytechnique
PALAISEAU, FRANCE

Dear Members of the École Polytechnique Admissions Committee,

The Polytechnique tradition — the belief that the most pressing human challenges require the most rigorous scientific training — is one I encountered not through a guidebook but through reading. The alumni list reads like a map of modernity: engineers, mathematicians, physicists who reshaped the world. I do not presume to follow in those footsteps; I presume only to share the aspiration.

The Bachelor of Science at l'X is, to my knowledge, one of the few undergraduate programmes anywhere that treats scientific breadth not as a compromise but as a philosophy — the conviction that a researcher in any discipline is a better researcher for having deeply studied several. I share that conviction entirely.

My academic trajectory has been shaped by a similar philosophy. In Vietnam, I pursued mathematics, informatics, physics, and literature simultaneously — not because the curriculum required it, but because I believe each discipline sharpens the others. I ranked first nationally in my cohort, earned a Gold Medal in the Mathematical Olympiad, and built a research project on Vietnamese medical AI in the gaps between these commitments.

I am learning French with seriousness and pleasure. The intellectual culture of Paris — the seminars, the debates, the refusal to treat any idea as beneath examination — is something I have long admired from a distance. I am eager to experience it firsthand, and to contribute to it in whatever way I am capable.

Je vous adresse mes respectueuses salutations,

Michael Nam

02 —

Curriculum Vitae

Education
Hanoi-Amsterdam High School
Specialization: Mathematics & Informatics
GPA: 3.98 / 4.0 · Class rank: 1st of 240 students
Stanford / Coursera — ML Specialization
Andrew Ng · 98% aggregate score
IELTS Academic — 8.0 / 9.0
Awards & Honours
Gold Medal — VMO
Vietnamese Mathematical Olympiad
Top 10 National — VOI
Vietnamese Olympiad in Informatics
1st Place — HackHanoi AI Track
National Scholarship
Ministry of Education — Academic Excellence
Technical Skills
PythonExpert
Mathematics (Linear Algebra, Calculus)Advanced
Machine Learning (PyTorch, TensorFlow)Advanced
C / C++Proficient
Probability & StatisticsProficient
Git / GitHubProficient
Experience & Activities
Research Intern
VinAI Research Lab, Hanoi
Data pipeline development for NLP corpus; literature review on Vietnamese language models.
Co-founder, AI Club
Hanoi-Amsterdam HS · 45 members
Volunteer Math Tutor
Underprivileged students, rural Hanoi

03 —

Academic Transcripts

Cumulative GPA
3.98
Rank 1 of 240 · Class of 2025
Hanoi-Amsterdam High School — Specialization in Mathematics & Informatics. Three consecutive years of academic distinction awards. National scholarship recipient from the Vietnamese Ministry of Education.
Certified transcript available upon request.
SubjectYearScoreGradeNote
Advanced Mathematics 2022–2025 9.8 / 10 A+ Top of cohort
Informatics / Computer Science 2022–2025 9.9 / 10 A+ Top of cohort
Physics 2022–2025 9.5 / 10 A
English Language 2022–2025 9.7 / 10 A+ IELTS 8.0
Chemistry 2022–2025 9.2 / 10 A
Literature & Humanities 2022–2025 9.0 / 10 A
Cumulative Average 9.6 / 10 A+ GPA 3.98 / 4.0

04 —

Letters of Recommendation

Michael is, without reservation, the most mathematically gifted student I have taught in twenty-two years. His solutions to competition problems regularly reveal approaches I had not considered myself. Beyond raw talent, he possesses the rare quality of intellectual humility — he is the first to say he does not yet understand something, and the last to leave until he does.
N
Dr. Nguyễn Thị Lan
Head of Mathematics · Hanoi-Amsterdam HS · 22 years teaching
During his internship at VinAI, Michael built a data preprocessing pipeline in two weeks that our team had been postponing for months. What struck me was not just his technical speed, but his instinct to ask the right questions first. He is already thinking like a researcher, not a student.
T
Dr. Trần Văn Minh
Research Scientist · VinAI Research Lab · PhD, University of Edinburgh
As his CS teacher, I have watched Michael go from student to peer. He has consistently helped classmates understand concepts I struggled to explain, and his AI Club has become one of the most active student organizations in the school's history. He leads by example and elevates everyone around him.
H
Mr. Hoàng Đức Bình
Computer Science Teacher & Department Chair · Hanoi-Amsterdam HS

05 —

Research Proposal

AI for HealthNLPLow-Resource LanguagesSoutheast AsiaFairness in ML
Towards Equitable AI Diagnostics in Low-Resource Clinical Settings: A Vietnamese Language Baseline

Motivation. The overwhelming majority of clinical AI research is conducted in high-resource languages — English, Mandarin, German — using datasets that structurally exclude populations in Southeast Asia. Vietnamese, spoken by over 95 million people, has fewer than a dozen openly available medical NLP corpora.

Proposed Research Direction. I propose to investigate transfer learning methods — specifically multilingual pre-trained transformers (mBERT, XLM-R) — for Vietnamese medical entity recognition and symptom classification, asking how much labeled data is required and how demographic biases propagate across language boundaries.

Preliminary Work. A dengue symptom classifier trained on 1,200 manually annotated Vietnamese patient summaries achieves 87.3% F1 — encouraging for its simplicity, but clearly insufficient at scale. This project revealed both the promise and the severe limitations of working without adequate data infrastructure.

Fit with École Polytechnique. Mathematical foundations of machine learning — generalization theory, PAC learning, and their implications for model reliability in high-stakes medical AI applications.


06 —

GitHub Projects

github.com/michaelnam-ai · 12 public repositories · 340+ contributions in 2024
Python
dengue-classifier
Symptom-based dengue fever classifier trained on 1,200 Vietnamese patient records. 87.3% F1. Logistic regression + TF-IDF features.
★ 48⑂ 12
Python
nn-from-scratch
Fully manual neural network using only NumPy — no ML frameworks. Includes backprop, dropout, batch norm.
★ 112⑂ 34
Python
viet-nlp-bench
Benchmark comparing XLM-R, PhoBERT, and mBERT on Vietnamese NER tasks. Includes annotation guidelines.
★ 67⑂ 21
C++
competitive-algorithms
500+ solutions to competitive programming problems (Codeforces, USACO). Organized by topic and difficulty.
★ 89⑂ 27
Python
air-quality-lstm
LSTM model predicting Hanoi air quality (PM2.5) 24h ahead using historical AQICN data. RMSE: 14.2 µg/m³.
★ 55⑂ 18
JavaScript
ml-visualizer
Interactive browser-based visualization of gradient descent, decision boundaries, and neural network training. Used by AI Club.
★ 143⑂ 46