I’m an undergraduate student at Sukkur IBA University, obsessed with the mathematical heart of AI and how to ship it as useful software. I build tools and experiments that translate probability, optimization, and transformer internals into reproducible systems—think research notes that actually run.
In short: I study the math, build the models, and automate the boring parts so the interesting parts are repeatable.
A curated showcase of hands‑on work — with expandable details for quick scanning or deep dives.
🧑🏫 AI Study Coach
Repo: AI-Study-Coach
An AI-powered study assistant that helps generate structured learning plans and study workflows tailored to your goals. Built fully in Python for simplicity and reproducibility.
Key Features:
- Personalized study roadmaps and milestones
- Topic decomposition with suggested resources
- Modular planning for consistent, trackable progress
📌 Outcome:
Makes self-learning actionable by turning broad goals into clear, stepwise study plans.
Tech Used:
📄 Multiple‑PDF RAG Chatbot
Repo: Multiple-PDF-RAG-Chatbot
A simple RAG‑based chatbot that lets you ask questions over multiple PDFs and get responses grounded in your documents.
Key Features:
- Ingest and query multiple PDFs
- Chunking + embeddings + retrieval augmented generation
- Document‑aware answers for focused Q&A
📌 Outcome:
Turns static PDFs into an interactive knowledge base for faster information retrieval.
Tech Used:
💬 Sentiment Analysis using LSTM
Repo: Sentiment-Analysis-using-LSTM
A high‑impact sentiment analysis pipeline built with Bi‑Directional LSTM and a self‑attention mechanism.
Key Features:
- Bi‑Directional LSTM encoder with self‑attention
- End‑to‑end pipeline: preprocessing → training → evaluation
- Standard metrics (e.g., accuracy, precision, recall, F1)
📌 Outcome:
Delivers robust sentiment classification with a focus on sequence modeling quality.
Tech Used:
Bachelor of Engineering in CSE
Sukkur IBA University, Pakistan (Expected 2027)
- Deep dive into Probability & Statistics: Bayes' theorem, distributions, and concentration bounds.
- Experimenting with transformer internals: attention maps, positional encodings (RoPE/ALiBi).
- Building generative agents and tool-using models, focusing on planning and modular tool invocation.
Whether you want to talk about AI, collaborate on a project, or just say hi,
I'm always open to meaningful conversations and new opportunities!
⚡ Fueled by curiosity, built with code — Hamadullah