Skip to content
View smebad's full-sized avatar
:octocat:
Grinding 📚🎮❤️
:octocat:
Grinding 📚🎮❤️

Highlights

  • Pro

Block or report smebad

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
smebad/README.md

Hi, I'm Ebad 👋

Agentic AI Engineer • Machine Learning Engineer • LLM & RAG Systems


About Me

  • Building production-ready AI systems with a focus on LLMs, RAG pipelines, and agentic applications.
  • Building practical AI products with Python, PyTorch, Hugging Face, FastAPI, Streamlit, and SQL.
  • Interested in GenAI, AI agents, fine-tuning, and production deployment.
  • Practicing DSA and strengthening fundamentals through from-scratch implementations.

What I Build

  • RAG pipelines — vector retrieval, embeddings, grounded LLM responses
  • LLM-powered web apps — FastAPI + Streamlit, deployed and production-accessible
  • Transformer systems — implemented from scratch in PyTorch, not just via APIs
  • Fine-tuned models — LoRA, PEFT, domain-specific datasets using Hugging Face
  • Agentic AI assistants — multi-turn context, tool use, goal-directed interactions

Current Work

🏠 RealtyIQ View Project

An end-to-end AI platform for property price prediction, semantic search, and conversational listing intelligence:

  • Price Prediction — XGBoost regression model with SHAP explainability
  • Semantic Search — Embedding-based property search using Sentence Transformers + FAISS
  • AI Assistant — RAG-powered chatbot grounded in real listing data (Hugging Face LLM)
  • REST API — FastAPI backend with full CRUD and inference endpoints
  • Dashboard — Streamlit UI for predictions, search, and chat

Featured Projects

🧩 Transformer From Scratch

Complete Transformer architecture implementation in PyTorch, inspired by Attention Is All You Need.
Built with modular code, custom tokenizer, and sequence modeling training pipeline.
View Project

🤖 PregniSense

AI-powered pregnancy risk assistant built with LlamaIndex + Gemini + Streamlit.
Uses retrieval-grounded responses and structured risk classification.
View Project

📚 Smart Study Buddy

LLM-powered document Q&A assistant built with RAG, embeddings, LangChain, Groq, and ChromaDB.
Answers questions from uploaded study material intelligently.
View Project

💻 Code Explainer

FastAPI + LLM app that explains, debugs, and improves code in multiple languages.
Deployed on Railway.
View Project

🪄 AI Portfolio Generator

Streamlit app that generates structured portfolio content from user input using Groq's Llama 3.3-70B.
View Project

🧠 Fine-Tuning Models

Experimentation repo for open-source LLM fine-tuning, training workflows, and evaluation.
View Project


Tech Stack

AI / ML

PyTorch TensorFlow Scikit-Learn Hugging Face

GenAI / LLM

LangChain LlamaIndex Streamlit

Languages

Python JavaScript C++ SQL

Data / Tools

NumPy Pandas Matplotlib Plotly OpenCV

Backend / Deployment

FastAPI Flask Docker MySQL


GitHub Stats


Connect With Me

Pinned Loading

  1. Transformer-from-Scratch Transformer-from-Scratch Public

    This project is a complete, from-scratch implementation of the Transformer architecture as described in the "Attention is All You Need" paper.

    Python

  2. PregniSense PregniSense Public

    AI-powered pregnancy risk assessment app that classifies pregnancy symptom risks using a medical knowledge base using Google Gemini + RAG + Streamlit

    Python 1

  3. AI-portfolio-generator AI-portfolio-generator Public

    AI Portfolio Generator is a Streamlit-based web application that leverages Groq’s Llama 3.3-70B model to generate fully responsive, professionally structured portfolio websites in seconds. It autom…

    Python 1

  4. Fine-Tuning-Models Fine-Tuning-Models Public

    In this repository I will be fine-tuning the different open source models and will test them out on the different datasets

    Jupyter Notebook 1

  5. Code-Explainer Code-Explainer Public

    An AI-powered web app that instantly explains, debugs, and improves any code you paste. Built with FastAPI and Groq's Llama 3.3-70B, supports Python, JavaScript, SQL and more. Deployed live on Rail…

    HTML 1

  6. Smart-Study-Buddy Smart-Study-Buddy Public

    AI-powered PDF Q&A chatbot. Upload any document and have a real conversation with it. Built with RAG architecture using LangChain, Groq (Llama 3.3-70B), ChromaDB, and HuggingFace embeddings, comple…

    Python 1