AI ENGINEER

MOHAMED KHALED_

AI & ML COMPUTER VISION DATA ANALYSIS DATA SCIENCE

Turning ideas into intelligent solutions with the power of AI.

Mohamed Khaled — AI Engineer

About Me

Artificial Intelligence graduate (2026) specializing in Generative AI, Retrieval-Augmented Generation (RAG), Agentic AI, Computer Vision, and Data Analysis.

Experienced in building production-ready LLM applications, multi-agent architectures, real-time computer vision systems, and AI-powered assistive technology using Python, LangChain, and the Gemini API. Currently advancing data science skills through the IBM Data Scientist track (DEPI). Holds 10+ industry certifications spanning AI, data science, and software development.

0

AI Projects Built

10+

Certifications

IBM

Data Scientist Track

NEURAL NETWORK ARCHITECTURE

Technical Skills Topology

Interactive Neural Synapses — Hover over any node to highlight synaptic connections across foundational, deep learning, vision, and LLM layers.

INPUT · FOUNDATIONS
Python (Advanced)
Git & GitHub
SQL Databases
MS Excel (Advanced)
LAYER 2 · DATA SCIENCE
Pandas & NumPy
Data Analysis & Viz
Model Evaluation Metrics
LAYER 3 · DEEP LEARNING
Scikit-Learn (ML)
Transformers & Cross-Encoders
LAYER 4 · GENAI & AGENTS
LangChain & Prompt Eng
Vector DBs (ChromaDB)
OUTPUT · VISION & APPS
Gemini API & Multimodal
FastAPI & REST APIs
AI PRODUCTION CAPABILITIES

Engineering Services

Enterprise-grade artificial intelligence solutions engineered for precision, speed, and real-world impact.

PRODUCTION LLM

Retrieval-Augmented Generation (RAG)

Custom enterprise RAG systems eliminating hallucination. Built with hybrid vector-lexical search (ChromaDB + BM25/TF-IDF), multi-query expansion, cross-encoder reranking, and automated 6-metric response quality evaluation.

LangChain Gemini API ChromaDB Cross-Encoder
AUTONOMOUS

Agentic AI & Multi-Agent Workflows

Autonomous AI agents equipped with planning, tool calling, memory management, and self-correction loops. Orchestrated with LangGraph to automate complex multi-step workflows, research, and data operations.

LangGraph Multi-Agent Tool Calling Autonomous
EMBEDDED & EDGE

Computer Vision & Assistive Systems

Real-time visual perception pipelines from edge cameras to cloud APIs. Specializing in YOLOv8 real-time object detection, face recognition, OCR document analysis, and wearable assistive technologies (e.g. Smart Glasses).

YOLOv8 OpenCV Edge AI Face Recognition
DEPI · IBM TRACK

Data Science & Predictive Analytics

Comprehensive data pipelines from exploratory data analysis and feature engineering to statistical hypothesis testing, classification/regression modeling, and interactive dashboards.

Python Pandas Scikit-Learn IBM Track
PORTFOLIO INNOVATIONS

Selected Work

AI engineering projects — from a real-world graduation wearable to autonomous cybersecurity agents.

FLAGSHIP GRADUATION PROJECT

Smart Glasses

An AI-powered wearable assistive system for the visually impaired. Integrates real-time YOLOv8 Object Detection, Gemini OCR, Face Recognition, and full Arabic Voice Control — enabling independent navigation and daily interaction.

30+ FPS Edge Vision YOLOv8 Gemini AI Arabic Voice Emergency SOS Python OpenCV Currency ID
Smart Glasses AI — wearable assistive device
FEATURED PROJECT

More Projects

Cognitive Hybrid RAG System
RAG ENGINE

Cognitive Hybrid RAG

Production-ready hybrid RAG system combining ChromaDB vector search + TF-IDF lexical retrieval with Cross-Encoder reranking and LLM query expansion — delivering hallucination-free answers.

PythonChromaDBLangChainGemini APIStreamlit
Sentinel-AI — Multi-Agent Fraud Detection
CYBERSECURITY

Sentinel-AI

Intelligent multi-agent fraud detection system with three specialized AI agents — Analyzer, Forensics, and Decision — providing real-time transaction security verdicts.

PythonGemini AIStreamlitMulti-Agent
AUT Smart Chatbot
CHATBOT

AUT Smart Chatbot

Bilingual Arabic/English AI assistant for Aqaba University of Technology. Built with Gemini 2.0 Flash, featuring GPA calculation, faculty directory, and a custom NLP pipeline.

JavaScriptGemini 2.0NLPHTML/CSS
PixelForge — Real-Time Image Filtering Studio
IMAGE PROCESSING

PixelForge

Real-time image filtering web app with 20+ filters, 8 one-click presets, live histogram, and multi-format export — built entirely in vanilla HTML5 Canvas with zero dependencies.

HTML5Canvas APIVanilla JSCSS3

Training & Certifications

ISO Artificial Intelligence Training

MediX International Accredited Institution

Applications of Artificial Intelligence (15 Hours)

The British University in Egypt (SCE)

Artificial Intelligence Training

Telecom Egypt (WE)

Python & AI Course (36 Hours)

Black Horse Courses — Grade: Excellent

Integrated English Program (94%)

Egyptian Foundation for Information Technology

View Official Certificates
OPEN FOR INQUIRIES

Get In Touch

Send a direct message or reach out through any of my communication channels below.