Building intelligence from scratch — one transformer at a time.
I'm a Computer Science & Engineering student at Daffodil International University (2025–2028), with a deep passion for Artificial Intelligence — particularly Large Language Models (LLMs).
My journey started with Python and data analysis, quickly evolving into local LLM deployment using Ollama, studying transformer architectures, and experimenting with model fine-tuning. I'm an active researcher in two university clubs, building the theoretical and practical foundations for my long-term ambition: building my own LLM from scratch before graduation.
I believe in learning by doing — my projects are proof of that philosophy, bridging theory with real-world implementation.
A context-aware AI companion leveraging local LLM deployment via Ollama. Features a custom memory management system with user profiling, conversation history, task/reminder tracking, and adaptive personality architecture with mood-based dynamic responses.
github.com/itzRatul/M.E.R.C.Y →
Processed 24 years (2001–2024) of daily rainfall data from multiple geographic locations. Built automated data pipelines, applied linear regression for forecasting, and generated comprehensive climate visualizations including trend lines and seasonal patterns.
github.com/itzRatul/project-rainfall →
A secure command-line healthcare record management system with multi-user authentication, comprehensive patient profile management, intelligent search (by name, contact, disease, blood group), and persistent file storage. Built entirely in C.
github.com/itzRatul/Patient-management-system →
Welcome to my digital garden. A space where I document my explorations in building Large Language Models, decode complex AI architectures, and reflect on the philosophy of continuous learning. It's more than just tutorials—it's a transparent window into my journey from a curious CS student to an AI researcher.
Open to research collaborations, internship opportunities, and AI discussions.