Tech Stack
Description
In the 'Review My Article' project, I collaborated with a client from Fiverr to develop a platform capable of generating AI-based summaries of PDF documents. This innovative tool enables users to submit PDF documents and receive an organized summary, categorized into Context, Key Insights, Key Findings, and Definitions. Not only does this allow a user to grasp the essence of an article or document quickly, but it also enhances comprehension by presenting the summary visually.
The platform features a PDF viewer, implemented using the PDF.js library, alongside summary cards for each category. This dual-pane layout facilitates a deeper understanding of the document by allowing users to view the original document and the summary simultaneously.
Under the hood, I utilized Python and the Summarization Machine Learning model from Huggingface's Transformers library to analyze and generate the summaries. To expedite the analysis process, I employed a Python Multiprocessing approach, dividing the document into chunks and processing them in parallel across all available server cores.
This project showcases my ability to integrate cutting-edge machine learning models into practical applications, optimize computational resources, and create user-friendly interfaces that enhance user experience and understanding.