> Initializing AI System...
> Loading Mohamed Abdelbasit...
> Building Intelligence Engine...
> System Ready.
> _
AI Engineer | AI Instructor
|
I'm Mohamed — an AI Engineer who delivers production-ready solutions.
I build AI that works — clean code, tested models, deployed systems. NASA Space Apps 1st Place, Data Pill Hackathon 3rd Place, and a graduation project graded A+. Today I ship production AI for enterprise clients across Saudi Arabia and the Gulf, with a focus on Arabic-language AI and real-world deployment. No experiments that die in notebooks — I ship.
Ready to build something that matters? Let's talk.
Faculty of Artificial Intelligence, Menoufia University
Graduation Project: Grade A+ — team leader of a 7-member team building a multi-model factory monitoring system.
AMIT Learning Academy
Thakaa Flow · Saudi-based technology company
Delivering production AI systems for large Saudi and Gulf enterprise clients — owning the full lifecycle from problem framing and data pipelines through model training, API development, and deployment. Building Arabic conversational and voice AI agents, leading an edge-AI wearable for blind users on NVIDIA Jetson Orin Nano (scene description, OCR, currency recognition, obstacle alerts), and shipping an internal enterprise platform end to end with Next.js 15, Supabase, and Vercel.
Digital Egypt Pioneers Initiative (DEPI) · Internship
Designed and implemented ML pipelines. Worked with production data, model training, evaluation, and deployment. Collaborated with engineers on real-world AI solutions.
Skills Area
Taught AI and Data Science to 150+ students — Python, data analysis, supervised and unsupervised learning, and model evaluation. Delivered hands-on workshops and technical sessions, and mentored students through end-to-end projects from data preparation to result interpretation.
Independent
Delivering end-to-end AI solutions for clients: predictive models, computer vision, NLP, and RAG systems. From data pipelines and model training to deployment and documentation — shipped and maintained.
You need results — not experiments that never ship.
NASA Space Apps 1st Place. Data Pill 3rd Place. 5 production projects. I deliver.
I explain complex AI in simple terms. No jargon. You'll always know the status.
From your data to a working system. I handle the full pipeline.
Realistic timelines. Transparent process. You get what we agree on.
Most clients need one or more of these. I build them all.
Forecast sales, churn, risk — turn historical data into future insights.
Image classification, object detection, face recognition, quality inspection.
Text analysis, sentiment, extraction, automated responses.
Personalize products, content, or suggestions for your users.
Spot fraud, failures, or outliers in real-time data streams.
APIs, web apps, dashboards — get your model into production.
Automate data ingestion, cleaning, and transformation for ML readiness.
Custom chatbots, document Q&A, and intelligent agents with your data.
Digitizing the informal resale market for used electronics in the Arab region and improving trust in second-hand transactions.
Computer Vision, Generative AI, RAG · Machine Learning, Prompt Engineering
AI-powered marketplace that evaluates device condition, estimates fair regional pricing, and helps retailers reduce electronic waste while increasing profitability.
Textile factories inspect fabric quality, safety compliance, and attendance manually — slow, inconsistent, and impossible to audit at scale.
YOLOv8, STPM Anomaly Detection, Facial Recognition · PyTorch, OpenCV, FastAPI, Arduino
🎓 Graduation Project — Grade A+ · Team Leader of 7
93% mAP@0.5 PPE safety detection, 90% AUROC fabric defect detection, and 96% facial-recognition attendance — unified in one dashboard with Arduino-controlled on-floor alerting.
Elderly people living alone face falls, fires, and silent health decline with no one watching — families need eyes on risk, not a call after the emergency.
YOLO, XGBoost, Emotion Recognition · PyTorch, OpenCV, FastAPI, Node.js, WebSocket
🥉 3rd Place — Data Pill Hackathon 2026 · Team Leader
Four production models: fire & smoke detection at 91.5% mAP@0.5, fall detection at 90.2%, 6-class emotion recognition at 87.3%, and a 99%-accuracy XGBoost health risk score — plus an Arabic voice companion.
Road hazards and damaged traffic signs go unreported for months because filing a formal incident report is slow, manual, and rarely reaches the right authority.
YOLO + MobileNet two-stage pipeline, Google Gemini · TensorFlow, n8n, React, Vite
Detects road hazards and auto-generates formal Arabic incident reports routed to the responsible authority — with an RTL authority dashboard and a citizen portal showing live status timelines.
I don't just build AI — I teach it. Hands-on workshops and full courses that take students from zero to shipping their own models.
Full Course
A complete program taking students from Python fundamentals to trained, evaluated models. Data analysis, supervised and unsupervised learning, model evaluation, and end-to-end projects — every concept paired with hands-on practice.
Workshop
What an AI engineer actually does day to day — the difference between a notebook experiment and a deployed system, and the pipeline that connects them. Full session recording available.
Workshop
The honest path into AI engineering — what to learn, in what order, and what to skip. Built from real hiring expectations, not tutorial lists.
Workshop
Practical AI for students — using LLMs to understand material faster, build study plans, and check their own work, without outsourcing the thinking. Delivered online in partnership with Obour Team.
Giving back to the community while growing my skills.
Shabab Betesaed Shabab
Guiding students through AI & ML fundamentals. Hands-on mentorship and project reviews.
Compiler Community
Contributing to innovative projects. Leadership, collaboration, and technical growth.
Shabab Betesaed Shabab
Mentoring youth in programming, marketing, design. Organizing large-scale events.