EPOCH SOCIETY PRESENTS

Collaborated with

16 HOURS HACKATHON

Event Date: November 5th & 6th

Event Start Time:9:30 AM

Venue:APJ Abdul Kalam Lab 4th floor, Sir C V Raman Block

Prize Pool: ₹10,500

Organized by: EPOCH SOCIETY

Key Topics

• Learn cutting-edge Gen AI concepts from SmartBridge experts
• Work with industry professionals
• Earn certificates & win exciting prizes
• Network with innovators, developers & tech leaders

PROBLEM STATEMENTS

AI Medical Prescription Verification Leveraging IBM Watson and Hugging Face Models ↴
PROJECT OVERVIEW
The Medical Prescription Verification System is an AI-driven healthcare project designed to ensure safe and accurate prescriptions using IBM Watson and Hugging Face NLP models. It automatically detects drug interactions, verifies dosages, and recommends safer alternatives based on patient details. The system is built using Python with FastAPI for the backend and Streamlit for a real-time, interactive frontend. To build it, you’ll use IBM Watson API for medical language understanding, Hugging Face Transformers for NLP-based text extraction, Pandas and NumPy for data processing, SQLite or MongoDB for storing patient and drug data, and REST APIs for integration between modules. Additionally, matplotlib or Plotly can be used for visual analytics, and scikit-learn or TensorFlow may be included for dosage prediction or recommendation models. This complete setup ensures an intelligent, accurate, and user-friendly prescription verification system.
ClauseWise: Legal Document Analyzer Using IBM Watson & Granite ↴
The ClauseWise: Legal Document Analyzer is an AI-powered tool that simplifies, decodes, and classifies complex legal texts to make them easily understandable for lawyers, businesses, and general users. It leverages IBM Watson, Granite, and Hugging Face models to automate the interpretation of legal clauses, obligations, and document types, reducing the manual effort involved in legal review. The system performs clause simplification by rewriting complex legal language into clear, plain text and uses Named Entity Recognition (NER) to extract important details such as parties, dates, amounts, and obligations. It also enables clause extraction and breakdown from lengthy documents for deeper analysis and provides document type classification to categorize files like NDAs, leases, employment contracts, and service agreements. Supporting multiple file formats including PDF, DOCX, and TXT, it ensures flexible document uploads. Developed using Python, the tool features an intuitive and interactive interface built with Streamlit and Gradio, making legal document analysis faster, smarter, and more accessible with the help of advanced AI technologies.
Personal Finance Chatbot: Intelligent Guidance for Savings, Taxes, and Investments ↴
The Personal Finance Chatbot: Intelligent Guidance for Savings, Taxes, and Investments is an AI-powered chatbot that provides personalized financial advice using IBM Watson and Generative AI models. It offers insights on savings, taxes, investments, and spending habits, adapting its tone and complexity based on the user type, whether students or professionals. The chatbot helps users make informed financial decisions while improving their overall financial literacy. Key features include personalized financial guidance for savings and investments, AI-generated budget summaries that present spending data in an easy-to-understand format, and spending insights that suggest better expense management strategies. It also uses demographic-aware communication to tailor its responses and a conversational NLP experience for natural, context-aware financial Q&A powered by IBM’s NLP tools. Developed using Python, Streamlit, IBM Watson, Hugging Face, and Granite, this intelligent chatbot combines technology and finance to deliver accessible, real-time financial guidance for users of all backgrounds.
StudyMate: An AI-Powered PDF-Based Q&A System for Students ↴
The StudyMate: An AI-Powered PDF-Based Q&A System for Students is an AI academic assistant designed to help students interactively learn from their study materials. It enables users to upload PDFs such as notes, textbooks, or research papers and ask questions directly, receiving precise, contextual answers extracted from the uploaded content. Using advanced NLP and embedding-based search, the system eliminates the need to manually search or read through lengthy documents. Key features include conversational Q&A for natural-language queries, text extraction and preprocessing powered by PyMuPDF, and semantic search capabilities implemented with FAISS and embeddings. It generates accurate, grounded answers through LLM-based processing using IBM Watson’s Mistral model. Built with Python and a Streamlit-based user interface, StudyMate ensures a seamless experience for document uploads, question handling, and intelligent response generation. The system integrates technologies like Hugging Face, PyMuPDF, Mistral, and FAISS to create an efficient, interactive, and student-friendly learning companion.
EchoVerse – An AI-Powered Audiobook Creation Tool ↴
The EchoVerse – An AI-Powered Audiobook Creation Tool is a generative AI-based system that transforms written text into expressive, natural-sounding audio. Designed to enhance accessibility and content reusability, it benefits students, content creators, and visually impaired users by providing an easy way to listen to written material. The tool allows customization of tone — such as Neutral, Suspenseful, or Inspiring — and supports different voice styles for personalized narration. Key features include tone-adaptive text rewriting powered by IBM Watson Granite, high-quality voice narration using advanced Text-to-Speech (TTS) technology, and downloadable audio output in .mp3 format for offline access. It also offers a side-by-side comparison of original and rewritten text to enhance user control and editing precision. Built with Python and featuring a Streamlit-based interface, EchoVerse provides a seamless, interactive, and user-friendly experience for transforming any text into dynamic audio content.

"The future is just one epoch away"