Jean Youakim, Computer Engineering student
Jean Youakim
I build machine learning systems and hold them to real-world results.
Computer Engineering at LAU, CGPA 3.99. My wildfire detector flags 97.8% of fire and smoke images, my World Cup model called all four semifinalists before kickoff, and I'm now building a red-team arena that teaches LLM agents to resist prompt injection.
Prefer scrolling? Everything in the terminal is laid out below.
About
I'm a third-year Computer Engineering student at LAU in Byblos, working mostly in applied machine learning: computer vision, LLM agents and their security, and predictive models.
This summer I finished the LebNet Tech Fellows AI track in the top third, with a Certificate of Distinction (33 of 102 fellows). My capstone, a wildfire early-warning system, catches 97.8% of fire and smoke images, and 98.2% of its alerts are real.
I also like knowing what happens below the software. In team lab projects I've helped build a 47-chip logic system on breadboards and an analog circuit that pulls a heartbeat out of a noisy sensor.
Outside engineering, I teach piano to four students, play futsal for LAU, and play chess.

- Studying
- BE in Computer Engineering, LAU
- Based in
- Lebanon
- Speaks
- Arabic (Native), English (C2), French (C1), Spanish (B1), German (A1)
- Looking for
- Open to software, AI/ML and engineering internships.
Experience
Newest first, like a commit log.
Jun 2026 – Aug 2026
AI Track Fellow
LebNet Tech Fellows Program
- Graduated with a Certificate of Distinction, awarded to 33 of 102 graduating fellows.
- Three-month curriculum spanning classical ML, neural networks, transformer architectures, LLM fine-tuning, RAG and agentic AI systems.
- Capstone: Wildfire Detection, a fire and smoke early-warning system.
2024 – Present
Private Piano Instructor
Self-employed
- Teach four private students with individually tailored curricula.
- Handle scheduling and lesson planning on my own.
Projects
Applied AI first, with the hardware I've built along the way.
AI / MLLebNet Fellowship capstoneAug 2026
Wildfire Detection
A fire and smoke early-warning system that turns camera, drone and phone images into a single alert.
Python, PyTorch, YOLOv8m, OpenCV, Streamlit, pytest
- alert recall
- 97.8%
- alert recall
- alert precision
- 98.2%
- alert precision
AI / MLPersonal projectJun 2026
World Cup 2026 Predictor
An Elo and Random Forest model that simulated the 48-team tournament tens of thousands of times and called Spain as champion before kickoff.
Python, pandas, NumPy, scikit-learn, Jupyter
- matches rated
- 15,790
- matches rated
- tournament simulations
- 50,000+
- tournament simulations
AI / MLCOE 548 course project, LAUFall 2026In progress
Prompt-Injection Arena
A red-versus-blue multi-agent system where an attacker tries to make an LLM email assistant leak a secret, and defenders learn new rules after every breach.
Python, LangGraph, LangChain, RAG, Chroma, Pydantic
HardwareLogic Design Lab, LAU2026
Smart Attendance System
A classroom attendance machine breadboarded from 47 74LS logic chips, with a countdown, a counter and a state machine.
Quartus Prime, 74LS logic ICs, JK flip-flops, FSM design
- logic ICs
- 47
- logic ICs
- countdown I built
- 10 → 0
- countdown I built
Smaller builds
HardwareElectronics I Lab, LAU
Heart Pulse Monitor
An analog circuit that pulls a clean heartbeat out of a noisy optical sensor with multi-stage op-amps and active band-pass filters.
Team project
SoftwareData Structures Lab, LAU
The Stacks Library
A web app for building, organising and searching a personal library of books.
Team project
Skills
Where a tool shows up in a project, the project is linked under it.
LLMs and agents
- LangGraphPrompt-Injection Arena
- LangChainPrompt-Injection Arena
- RAGPrompt-Injection Arena
- ChromaPrompt-Injection Arena
- PydanticPrompt-Injection Arena
- Prompt-injection defencePrompt-Injection Arena
Machine learning and vision
- PyTorchWildfire Detection
- YOLOv8Wildfire Detection
- OpenCVWildfire Detection
- scikit-learnWorld Cup 2026 Predictor
- pandasWorld Cup 2026 Predictor
- NumPyWorld Cup 2026 Predictor
- StreamlitWildfire Detection
Hardware and embedded
- Quartus PrimeSmart Attendance System
- 74LS logic ICsSmart Attendance System
- FSM designSmart Attendance System
- ProteusHeart Pulse Monitor
- Op-amp circuitsHeart Pulse Monitor
- HCS12 Assembly
- ROS 2 fundamentals
Languages
- PythonWildfire Detection, World Cup 2026 Predictor
- C++
- Java
- SQL
- JavaScriptThe Stacks Library
- HTML and CSSThe Stacks Library
Tools
- Git and GitHub
- Linux
- pytestWildfire Detection
- MongoDB
Education
Sep 2024 – Spring 2028 (expected)
BE in Computer Engineering
Lebanese American University, Byblos, Lebanon
- Cumulative GPA
- 3.99
- Cumulative GPA
- Major GPA
- 4.00
- Major GPA
Relevant coursework
AI
- Machine Learning
- Large Language Models
- Trustworthy & Secure AI
Software
- Data Structures & Algorithms
- Database Systems
Hardware
- Computer Architecture
- Microprocessors
- Digital Logic Design
- Signals & Systems
Certifications
- Machine Learning Specialization, DeepLearning.AI (Andrew Ng)
- Deep Learning Specialization, DeepLearning.AI (Andrew Ng)
Honours and clubs
- Ranked in the top 30 nationally in the Lebanese Baccalaureate, General Sciences track
- Member of the IEEE AI Club and the Software Engineering Club
Outside class
- Scouts de l'Indépendance for seven years (2018–2024), leading youth camps and civic initiatives
- Piano: Level 7A certification from the Antonine School of Music
- LAU futsal team, and a former U16 national futsal competitor
- 1500 rapid rating in chess
- Class representative on the student council in grade 12
Contact
Open to software, AI/ML and engineering internships. The fastest way to reach me is email.
jeanyouakim@gmail.com