ManuIndex is a document-aware retrieval engine for RAG that handles large, mixed document collections. Instead of searching all document chunks at once, it first finds the most relevant documents, then searches within the chunks of those documents. This approach improves retrieval efficiency and accuracy, especially for large datasets.

PyMuPDF4LLM is a Python library that extracts complex visual information (tables, charts, logos) from PDFs into Markdown using Vision Language Models. It bridges visual content and structured text with intelligent OCR, layout preservation, and multi-backend support (OpenAI, Hugging Face).

An automated movie ticket tracking application that monitors cinema websites for ticket availability and notifies users the moment tickets go on sale.

Academic profile template for scholars and designers. Powered by Eleventy & tsr-page theme.

A lightweight, real-time helmet detection

Fine-tuned distilbert model using LoRA on custom dataset

RAGmix is a heterogeneous, multi-domain evaluation dataset for Retrieval-Augmented Generation (RAG) systems.

A dataset of bank transactions labeled as fraudulent or non-fraudulent.

Compiled many small datasets into one cleaned dataset

Bash commands and installation guide for Fedora linux

All machine learning algos and case studies

All data structure and algos in single place