Home

Projects

Education

Experience

Skills

About

Connect

Graduate student in Machine Learning

Amir Bouraoui, MSc

Building intelligent systems and scalable ML solutions that drive business impact. Specializing in deep learning, computer vision, and production-ready AI applications.

See My Work

My Resume

Tools which I use on a daily basis.

Explore My Projects

Each project reflects a balance of creativity and practical impact, making it both original and effective.

Classical vs. Quantum Convolutional Neural Network Explainability for Brain Tumor MRI

This project compared a classical CNN with a hybrid Quantum CNN (QCNN) for binary brain tumor classification from MRI scans, evaluating performance and interpretability. Grad-CAM was used to visualize and follow how both models detected and reasoned about tumor regions.

View Project

Quantum Machine Learning

Explainability

Convolutional Neural Network

Python

Classical vs. Quantum Convolutional Neural Network Explainability for Brain Tumor MRI

This project compared a classical convolutional neural network with a hybrid Quantum CNN for binary brain tumor classification from MRI scans, evaluating both performance and interpretability. Grad-CAM was employed to trace and visualize the decision-making process of both models, allowing us to follow how each network detected and reasoned about tumor regions.

View Project

Classical vs. Quantum Convolutional Neural Network Explainability for Brain Tumor MRI

This project compared a classical convolutional neural network (CNN) with a hybrid Quantum CNN (QCNN) for binary brain tumor classification from MRI scans, evaluating both performance and interpretability. Grad-CAM was employed to trace and visualize the decision-making process of both models, allowing us to follow how each network detected and reasoned about tumor regions.

Quantum Machine Learning

Explainability

Convolutional Neural Network

Python

View Project

Local-LLM browser extension for Retrieval Augmented Search

PunaySearch is a privacy-first Chrome extension that runs a local AI to answer questions about any webpage. Using a Retrieval Augmented Generation (RAG) model powered by WebLLM and WebGPU, it provides contextual answers while ensuring no data ever leaves your device.

View Project

LLM

WebGPU

RAG

LangChain

Semantic Retrieval

Vector Embeddings

Local-LLM browser extension for Retrieval Augmented Search

PunaySearch is a privacy-first Chrome extension that runs a local AI to answer questions about any webpage. Using a Retrieval-Augmented Generation (RAG) model powered by WebLLM and WebGPU, it provides contextual answers while ensuring no data ever leaves your device.

View Project

Local-LLM browser extension for Retrieval Augmented Search

LLM

WebGPU

RAG

LangChain

Semantic Retrieval

Vector Embeddings

View Project

An Unbeatable Oware Bot through Self-Learning

Engineered a dominant AI for the strategy game Oware, designed to outperform its challengers. It uses a self-learning NegaMax algorithm, analyzing data from thousands of self-play simulations to dynamically adapt its strategy and achieve unparalleled performance.

View Project

Java

NegaMax

Heuristic Evaluation

Statistical Move Biasing

Alpha–Beta Pruning

An Unbeatable Oware Bot through Self-Learning

Developed an intelligent bot for the strategy game Awele using a NegaMax algorithm with alpha-beta pruning. The bot learns and improves by analyzing historical game data from self-play simulations, dynamically adjusting its strategy and search depth based on game phase and observed patterns to enhance decision-making.

View Project

An Unbeatable Oware Bot through Self-Learning

Java

NegaMax

Heuristic Evaluation

Statistical Move Biasing

Alpha–Beta Pruning

View Project

Multi‑Machine Task Scheduling Engine

I developed a scheduler in C using shortest processing time with both preemptive and non-preemptive modes, implemented core data structures including binary search trees, AVL trees, and linked lists, and connected the engine through the Java Native Interface.

View Project

Java

Greedy algorithms

Linked List

Multi‑machine scheduling

Multi‑Machine Task Scheduling Engine

Event‑driven scheduler with preemptive/non‑preemptive modes and optional balanced trees to minimize makespan across machines. Native C core exposed via JNI, with a JavaFX Gantt chart for visual results.

View Project

Multi‑Machine Task Scheduling Engine

Java

Greedy algorithms

Linked List

Multi‑machine scheduling

View Project

Show All

Education

Master's Degree in Machine Learning

View Details

University of Lorraine

Nancy, FRANCE

Bachelor's Degree in Computer Science

View Details

University of Lorraine

Metz, FRANCE

High School Diploma

View Details

Lycée des Arts et Métiers

Limpertsberg, LUXEMBOURG

Experience

Hands-on experience driving projects from concept to completion, with a focus on innovation, precision, and user-centered outcomes.

NICHE GUARDIaN | Luxembourg

NICHE GUARDIaN | Luxembourg

NICHE GUARDIaN | Luxembourg

Hired to spearhead the development of a real-time monitoring platform for managing a 200+ national kiosk network.

See Company Website

DELOITTE | Luxembourg

DELOITTE | Luxembourg

Recruited to design and implement intelligent automation solutions that streamlined critical departmental workflows.

See Company Website

Technical Skills

Python3

JavaScript

Java

C/C++

PyTorch

TensorFlow

Image Classification

NLP clustering

LLM Fine-tuning

LLM Deployment

LangChain

OpenCV

Docker

Git CI/CD

Proxmox

DevOps

LDAP

Agile methodologies

+ More

About me

Driven by a passion for

Artificial Intelligence

and grounded by a

Master’s in Machine Learning

, I develop

intelligent systems

that deliver

scalable solutions

for

real-world impact

Driven by a passion for

Artificial Intelligence

and grounded by a

Master’s in Machine Learning

, I develop

intelligent systems

that deliver

scalable solutions

for

real-world impact

Driven by a passion for

Artificial Intelligence

and grounded by a

Master’s in Machine Learning

, I develop

intelligent systems

that deliver

scalable solutions

for

real-world impact

Amir Bouraoui

Github