Top 10 / 67
LUXSAI Summit 2026 Hackathon
Placed in the Top 10 among 67 teams with Road Safety, a computer-vision system designed to anticipate road-accident risk before incidents occur, with FastAPI powering the backend.
Hi, I'm
Mohamed Alawakey
ML & AI Engineer
Junior AI Engineer building production-ready machine learning and LLM systems, with hands-on experience in model training and evaluation, dataset preprocessing, RAG pipelines with vector databases, secure API integrations, and scalable FastAPI backends. I focus on improving model performance and turning AI workflows into reliable deployed services.
Featured Projects
QualiVerse
Built a locally deployable document-management and accreditation-support system for universities, institutes, and schools. QualiVerse centralizes files, prevents local duplication without depending on an external provider, and uses AI models to generate course specifications and course reports—helping institutions pursue institutional and program accreditation while improving operational efficiency and support readiness. The system is deployed at the Faculty of Computers and Information, Tanta University, and is currently used by 20+ professors, teaching assistants, and Quality Assurance Unit staff.
PythonHugging FaceLLMsPrompt EngineeringFastAPI.NETFlutterRedisSQL Server
Partners
Faculty of Computers and Information, Tanta University
Deployment partner · Actively used by 20+ professors, teaching assistants, and Quality Assurance Unit staff.
RAGU
RAGU is an AI-powered educational platform that transforms documents and study materials into an interactive learning experience. It allows students to chat with their documents and receive accurate, source-based answers, extract text from scanned images, generate quizzes and flashcards, track their performance, and organize their studies using notes and mind maps. The platform supports Arabic and English, team collaboration, personalized learning tools, real-time responses, and subscription plans—all within one integrated study environment.
PythonFastAPINext.jsPostgreSQLMilvusRedisBM25Hugging Face EmbeddingsBGE RerankerLLMsPaddleOCRCeleryRabbitMQDockerNginxPrometheusGrafanaRAGAS
In Progress
AI Job Hunter System Agent
AI Job Hunter System Agent is an intelligent job-search assistant that helps candidates discover and evaluate relevant opportunities through an organized, automated workflow. It reads and structures the candidate’s CV, searches Wuzzuf based on their target roles, skills, and preferred locations, removes duplicate or unsuitable listings, and filters opportunities according to experience level, employment type, and work mode. It then analyzes each suitable job to present a clear summary of its responsibilities, required and preferred skills, experience expectations, useful keywords, and potential red flags—helping candidates focus their time on the most relevant opportunities while keeping them fully in control of the application process.
PythonLangGraphLangChainOllamaOpenAIPydanticpython-docxWeb ScrapingStructured LLM OutputAgentic Workflows
Paused
Hume AI
Hume AI is a bilingual AI writing assistant that transforms robotic or AI-generated text into writing that feels more natural, expressive, and personal while preserving its original meaning. It supports both Arabic and English, allows users to choose between standard, academic, creative, and casual writing styles, and processes texts of up to 3,000 words. The platform analyzes the text before and after rewriting, displays an AI-likelihood score to show the improvement, streams the result in real time, and includes convenient paste and copy tools within a responsive light and dark interface.
PythonFastAPICustom Linguistic Scoring AlgorithmServer-Sent Events (SSE)PydanticUvicornNext.js
Paused
Programming Book Recommendation System
Built a semantic book-recommendation platform with PyTorch and Sentence Transformers. Reduced a 10M+ record dataset to 400K+ curated entries, then implemented PGVector and HNSW search with sub-100ms latency, Redis caching, Docker, and automated GitHub Actions pipelines.
PyTorchSentence TransformersFastAPIPostgreSQLPGVectorRedisDockerCI/CD
Student Performance Predictor
Built a Scikit-learn model that predicts student performance with 91% accuracy. The project includes Pandas and NumPy preprocessing, feature engineering and exploratory analysis, a real-time FastAPI prediction service, Docker packaging, and deployment on Hugging Face Spaces.
Scikit-learnPandasNumPyFeature EngineeringFastAPIDocker
Professional Experience
GenAI EngineerGhaymah Systems
09/2025 – Present
Remote
- Designed and maintained MCP servers that connect LLM-powered applications with structured tools and external services.
- Built RAG pipelines with vector-database retrieval to ground generated responses in application knowledge.
- Developed Python REST APIs and secure integration flows, including controlled API-key handling, for production-oriented AI services.
Data Scientist TraineeALX Africa
01/2025 – 03/2026
Hybrid, Egypt
- Cleaned, analyzed, and preprocessed datasets with Python and SQL to turn raw data into actionable insights.
- Built and evaluated machine-learning models using validation workflows and performance metrics to compare model quality reliably.
- Applied the complete data-science workflow—from problem framing and preprocessing to model evaluation—across practical projects.
Artificial Intelligence InternNTI
04/2025 – 04/2025
Tanta, Egypt
- Applied core Machine Learning and Deep Learning algorithms through an intensive, hands-on AI training program.
- Explored Natural Language Processing techniques and connected them to practical use cases in intelligent systems.
Technical Skills
AI & Machine Learning
Machine LearningDeep LearningComputer VisionNLPLLMsPrompt EngineeringHugging FaceScikit-learnPyTorchLangGraphLangChainModel Optimization
Backend & APIs
PythonOOPFastAPIREST APIs.NETFlutterPostgreSQLSQL ServerRedisVector DatabasesMilvusPGVector
Data & Analytics
NumPyPandasMatplotlibSeabornData CleaningExploratory Data AnalysisData PreprocessingFeature EngineeringJupyter Notebook
DevOps & Tools
GitGitHubLinuxDockerDocker ComposeCI/CDGitHub ActionsVS CodePyCharm
Leadership & Working Style
Team leadership & teamwork
Led nine team members across Flutter, .NET, Data Analysis, Cybersecurity, AI, and Testing for the graduation project, while also serving as team leader across most university projects. Coordinated responsibilities, communication, technical dependencies, and shared decisions through deployment at the Faculty of Computers and Information, Tanta University.
Collaborative problem solving
Actively asks for different perspectives, discusses trade-offs, and compares multiple approaches to build a clearer understanding and reach stronger solutions.
Fast learning & growth
Learns new tools quickly, treats mistakes as feedback, and continuously improves both technical execution and communication.
Versatility, teaching & mentoring
Comfortable working across varied tasks. Teaching and mentoring experience strengthened communication, classroom management, and the ability to explain technical ideas clearly.
Languages
ArabicNative
EnglishProfessional (B2)
Achievements
Measurable project results, real deployments, competition recognition, and leadership milestones.
Selected from 5
University Deployment
QualiVerse was selected from five strong candidate projects for deployment at the Faculty of Computers and Information, Tanta University, while I was still an undergraduate. It is now used by 20+ professors, teaching assistants, and Quality Assurance Unit staff.
70–90%
Academic Workflow Time Saved
Automated course-quality workflows with two AI models: course specifications dropped from roughly 2–2.5 hours to at most 5 minutes, while course reports dropped from around 1–1.5 hours to about 30 seconds before minor faculty review and edits.
9 Members
Cross-functional Project Leadership
Led nine team members across Flutter, .NET, Data Analysis, Cybersecurity, AI, and Testing, coordinating a multidisciplinary graduation project from planning through university deployment.
91%
Prediction Accuracy
Achieved 91% accuracy with the Student Performance Predictor after preprocessing, feature engineering, and model evaluation.
View live project10M+ → 400K+
Dataset Curation
Cleaned and reduced more than 10 million book records into a curated dataset of over 400,000 useful entries.
View repositoryLet's Connect
I'm currently available for freelance projects and full-time roles in Machine Learning and AI. Feel free to reach out if you have a project in mind or just want to discuss the latest in AI.
Based in Cairo, Egypt