class MoazMohammed:
    def __init__(self):
        self.role = "AI Engineer in training"
        self.location = "Ismailia, Egypt"
        self.current_quest = "Understand the magic before importing the library"
        self.inventory = {
            "Python": "for teaching machines",
            "Go": "for understanding what the machines are standing on",
            "C++": "for when milliseconds start asking questions",
        }
        self.favorite_problems = [
            "search results that almost understand the question",
            "models that perform brilliantly until the test set arrives",
            "papers with one diagram and seventeen missing details",
            "APIs that deserve a database and a proper README",
        ]

    def learn(self, topic):
        return topic.read().implement().break_it().measure().document()

    def status(self):
        return {
            "curiosity": "unlimited",
            "experiments_running": True,
            "open_to_opportunities": True,
        }

I'm an AI Engineer and Computer Engineering student at Suez Canal University. I spend most of my time building things with machine learning, retrieval systems, computer vision, and LLMs, then figuring out how to make them actually work in production.

I also like getting my hands dirty with Go: HTTP servers, network protocols, CLIs, and database-backed APIs. Not because it's trendy, but because I want to understand the full stack my models sit on top of.

Currently looking for AI/ML, computer vision, or backend engineering internships and junior roles.

• Reproduced a CVPR 2016 group activity recognition architecture and pushed it to 88.86% accuracy (the paper reported 81.9%).
• Built a search and RAG engine from scratch, going through TF-IDF, BM25, semantic retrieval, reciprocal rank fusion, CLIP multimodal search, reranking, and cited generation.
• Shipped Modaresy, a tutor discovery product, in six weeks with a small team. Hit 60+ active users and placed 3rd at the Creativa Startup Competition.
• Built computer vision, sequence modeling, and agentic AI projects using PyTorch, TensorFlow, Hugging Face, LangChain, and LangGraph.
• Wrote a bunch of Go projects covering HTTP servers, TCP/UDP, PostgreSQL APIs, JWT auth, RSS aggregation, and CLI tools.

• My Setup: Neovim Config
• Go & systems: Linko · Peril · Tubely · HTTP From Scratch · Forge CLI · Pokédex · RSS Aggregator
• LLM work: Production RAG Engine · Applied LLM Engineering · LLM Mastery
• Applied ML: Fraud Detection · Facial Emotion Recognition · Transient Plotter · Asteroids

Machine learning: PyTorch · TensorFlow · Scikit-learn · Keras · OpenCV · XGBoost · LightGBM · CatBoost
LLM & retrieval: Hugging Face · LangChain · LangGraph · Groq · Gemini · RAG · BM25 · sentence-transformers · CLIP
Data: Pandas · NumPy · PostgreSQL · Matplotlib · Seaborn
Cloud & infra: AWS S3/CloudFront · RabbitMQ · Docker · OpenTelemetry · Prometheus
Apps & UI: Streamlit
Software: Python · Go · C++ · REST APIs · Git · Linux · Neovim

B.Sc. in Computer Engineering, Suez Canal University (expected 2028)
Coursework: machine learning, deep learning, computer vision, algorithms, linear algebra, probability, statistics.

• Turning experiments into tested, deployed services instead of just notebooks.
• Getting better at the production side: observability, migrations, failure handling.
• Building fewer projects but finishing them properly, with benchmarks, tests, and honest write-ups.
• Combining ML depth with Go and backend fundamentals so I can own a feature end to end.