I'm an AI Engineer passionate about building intelligent systems that combine language, vision, and reasoning. My work spans from fine-tuning LLMs and building AI agents to computer vision and classical ML applications — all with a focus on clean, scalable, and practical solutions.
- Build robust agentic solutions using suitable agentic patterns and designs
- Design and fine-tune Large Language Models for specialized tasks
- Build Retrieval-Augmented Generation (RAG) and AI Agent systems
- Develop computer vision models for detection and classification
- Engineer robust ML/DL pipelines using PyTorch, LangChain, FastAPI, and AWS
Languages: Python, SQL, C++, Java
Frameworks & Tools: PyTorch, LangChain, Langgraph, Pydantic, HuggingFace Tramnsformers, FastAPI, Scikit-learn, Streamlit, vLLM, Tensorflow
Databases: ChromaDB, Supabase, Postgresql, SQLite, MS SQL Server
Cloud: AWS (Cloud Architecture & ML Engineer), Docker
Other: Weights & Biases, Ollama,
FasAPI, Google GenAI SDK, Pydantic, ChromaDB An agentic workflow that utilizes a RAG pipeline to turn static pitch decks into an interactive live experience
LangChain, LangGraph, FastAPI, ChromaDB
A RAG-powered assistant for Obsidian that can summarize and answer questions based on your notes.
PyTorch, Weights & Biases
Built GPT-2 architecture from scratch and fine-tuned it for email spam classification.
Scikit-learn, Librosa, Streamlit
Developed a web app that predicts speaker gender from audio recordings using classical ML.
- LinkedIn: linkedin.com/in/hatemsal
- GitHub: github.com/HatemSal
- Email: hatemsalem2009@gmail.com