Get in touch

Hi, my name is

Reem Hussin Arafa.

AI & Data Science Engineer

I build intelligent systems — from LLM-powered tutors and grammar-correction pipelines to urban mobility research published at international venues.

📍 Giza, Egypt
View my workreemarafa2004@gmail.com

01About

I'm an Artificial Intelligence student at Nile University (GPA 3.7, President's Honor — ranked #1), specialising in machine learning, NLP, and generative AI. My work spans the full pipeline: framing the problem, wrangling the data, training and tuning models, and shipping something people can actually use.

I've presented urban mobility research at NetMob 2025 in Paris, published with IEEE, and interned at EVA Pharma applying LLMs to real pharmaceutical data. Alongside research, I've spent over two years as a teaching assistant — mentoring students, debugging their code, and learning how to explain hard ideas simply.

02Skills

Core AI

  • Machine Learning
  • Deep Learning
  • NLP
  • Generative AI
  • Computer Vision
  • Reinforcement Learning
  • Recommendation Systems
  • Prompt Engineering

Models & LLMs

  • Llama-3
  • GPT
  • BERT
  • T5
  • YOLO
  • CNNs

Frameworks & Libraries

  • Python
  • TensorFlow
  • Scikit-learn
  • Pandas
  • OpenCV
  • Whisper / Vosk
  • GECToR
  • CoEdit

Data & Analytics

  • Data Preprocessing
  • Statistical Analysis
  • Tableau
  • Power BI
  • Excel

MLOps & Tooling

  • MLflow
  • DVC
  • Git & GitHub
  • GitHub Codespaces

03Experience

  1. Jul 2025 – Aug 2025 · Cairo

    AI & Data Science Intern — Hackathon Track @ EVA Pharma

    • Selected from a national hackathon for a competitive internship focused on Generative AI, LLMs, and deep learning.
    • Applied advanced AI concepts to a hands-on project using real-world pharmaceutical data under mentor guidance.
    • Developed and presented a final data-driven solution to company stakeholders.

  2. Jul 2024 – Sep 2024 · Giza

    Research Intern — Data Analysis & Machine Learning @ Nile University

    • Spearheaded an end-to-end ML research project that resulted in a peer-reviewed paper submission with novel findings.
    • Built robust classification and prediction models, managing the full pipeline from preprocessing to hyperparameter optimisation.
    • Used Python (scikit-learn, Pandas) for deep statistical analysis, achieving significant improvements in model performance.

  3. Sep 2023 – Jan 2026 · Giza

    Junior Teaching Assistant — Computer Science @ Nile University

    • Supported professors in CS courses, reinforcing complex subject matter through lectures and personalised assistance.
    • Mentored students through lab sessions — guiding programming assignments, debugging code, and building problem-solving skills.

04Research & Publications

NetMob 2025 · Paris2025

Identifying First/Last Mile Deserts & Segmenting Human Mobility Behavior

Two co-authored research posters on urban mobility, presented at NetMob 2025 and published in the official Book of Abstracts.

IEEE ICCA2024

Deep Learning-Based Correlation Analysis of Public Transit and Ride-Sharing: Urban Mobility Patterns in Paris

Published in the IEEE International Conference on Computer and Applications (ICCA).

05Projects

VR + AI Tutor for Medical Education

An interactive VR anatomy learning experience with a fine-tuned GPT tutor built in.

  • Built a real-time VR anatomy experience in Unity for Meta Quest with organ interaction and voice navigation.
  • Integrated a fine-tuned GPT-based tutor using reinforcement learning to answer questions, generate adaptive quizzes, and personalise learning.
  • Unity
  • Meta Quest
  • GPT
  • Reinforcement Learning

Grammar Feedback for Language Learning

A grammar correction system reaching 96% accuracy by ensembling three models.

  • Engineered a correction pipeline combining T5, GECToR, and CoEdit models.
  • Added a BERT-based validation layer that flags grammatical errors and generates context-aware corrections.
  • T5
  • BERT
  • GECToR
  • NLP

Streaming Traffic Prediction — Paris & Marseille

Forecasting YouTube, Spotify, and Netflix traffic patterns in urban environments.

  • Predicted streaming traffic from historical datasets using machine learning models.
  • Evaluated accuracy via RMSE to extract insights for urban planning and network management.
  • Machine Learning
  • Time Series
  • RMSE
  • Urban Data

Real-Time Emotion Recognition

Multimodal emotion detection from face and voice, running in real time.

  • Built a CNN-based Mini-XCEPTION model for facial expressions and a voice model trained on RAVDESS and CREMA-D.
  • Fused vision and audio into a single responsive interface with OpenCV and TensorFlow.
  • CNN
  • OpenCV
  • TensorFlow
  • Multimodal

06Education

2022 – 2026 · Giza, Egypt

B.Sc. in Artificial Intelligence

Nile University — Faculty of Computer and Information Technology

GPA 3.7

  • 🏅 President's Honor — Spring 2024 (4.0 GPA, Ranked #1)
  • 🏅 Dean's Honor for sustained academic excellence — Fall 2025