AI Engineer with strong expertise in machine learning, deep learning, NLP, and computer vision. Experienced in designing and deploying production-ready AI solutions including chatbots, recommendation engines, and real-time detection systems. Skilled in full-stack development (MERN) and proficient in Python, C++, and JavaScript. Hands-on experience with cloud deployment (GCP, Azure) and MLOps practices. Passionate about applying AI to solve real-world problems and staying at the forefront of emerging AI technologies.
- 🎓 B.Sc. in
Computer Science & AI— Banha University - 💼 Currently work at COA Holding company
- 🧠 Always
learning new things& pushing boundaries - 🤖 Passionate about
Smart systems,AI,LLMs,Agents, andreal-world AI systems
|
Python |
PyTorch |
TensorFlow |
FastAPI |
Flask |
React |
Node.js |
Docker |
|
GCP |
Azure |
PostgreSQL |
MongoDB |
Neo4j |
Linux |
Git |
C++ |
Production systems, not just experiments.
🌍 Maika — AI Travel Assistant LangGraph RAG FastAPI GCP Azure
Personalized flight, hotel & destination recommendations with image-based location search. Multi-agent architecture hitting 10s average response time, with a Rust-powered web search tool that processes results in 3 seconds.
🏠 Deeb Realties — Cross-Platform AI Chatbot LangGraph Odoo GCP Facebook API
Deployed across WhatsApp, Instagram & Facebook Messenger with automated booking, complaint management, and full Odoo CRM integration. 99.9% uptime on GCP.
📢 Facebook Ads Optimization Agent LLM Agents GCP
Autonomous agent monitoring ad campaigns, adjusting budgets dynamically, and applying stop/continue logic based on predefined KPIs — eliminating manual intervention entirely.
🎙️ Real-Time Voice Agent with Avatar Gemini Live API WebSocket Google ADK
Voice-driven conversational agent with synchronized avatar rendered at 25 fps, achieving sub-2 second voice-to-animation latency via optimized WebSocket audio streaming.
🤝 Cooperative Multi-Agent System vLLM LangChain Docker GCP Azure
Feedback-loop architecture supporting GPT, Claude, Gemini and self-hosted local models. Extended with vision capabilities and live voice calling.
📝 Automatic Grading System LLM Vision LLM RAG FastAPI
LLM-powered grader achieving 94% accuracy on English and 86% on Arabic datasets — cutting manual grading time by 80%.
🔍 Real-Time AI Search Engine C++ gumbo curl Google Search API
End-to-end RAG pipeline with a high-concurrency C++ scraper. Processes 20+ websites and synthesizes answers in under 4 seconds — significantly faster than Python equivalents.
📊 CSV Report Builder LLM LangChain FastAPI
Natural language interface for CSV analysis — ask a question, get a full visual report with automated insight generation.
| 📦 Repository | 🔧 What I Did |
|---|---|
| ⚽ Tryolabs — Soccer Analytics | Extended YOLOv5/v8 support with dynamic label mapping, configurable team palettes & video params |
| 🤖 Mistral-Common | Standardized API responses, added streaming support, migrated to httpx for async |
| 🧠 MemoryOs | Fixed env variable bug preventing dynamic model assignment across utility functions |
| 📄 PageIndex | Added Docling support, Pydantic JSON validation & configurable LLM API |
I'm always open to discussing AI projects, research collaborations, or interesting engineering challenges.
Whether it's a production system or an ambitious idea — let's make it real.
| 📍 Location | 🌐 Kaggle | |
|---|---|---|
| emam200232@gmail.com | Cairo, Egypt 🇪🇬 | kaggle.com/elemam |



