R

Retrieval-Augmented Generation (RAG) for Scient...

Lifeline AI is an AI-powered emergency response platform providing real-time emergency intelligence, ambulance coordination, hospital connectivity, SOS management, and disaster response monitoring.

The central idea of the Telugu Sahitya Sahakaari project is to create a digital, community-powered archive for Telugu literature. The project's goal is to preserve literary heritage by allowing users to upload documents, images, and text. The users are also enabled to interact with an AI chatbot and explore, research, and learn about Telugu Literature.

✊ Swarajya Scanner — AI-Powered Legal Rights App in Your Language 🇮🇳 "Your rights. Your language. Your Swarajya."

Swarajya Scanner puts the Constitution in your pocket — and the confidence to use it. It's an AI-driven app that answers everyday legal questions in simple, regional language, helping citizens understand their rights without legal jargon.

Whether it's "Can police take my phone?" or "Do I need to give notice before quitting?", Swarajya Scanner breaks it down clearly, in languages people actually speak.

🌟 VakyaVibe.ai – Indian Language Corpus Collection Platform VakyaVibe.ai is an open-source, community-powered web application designed to preserve and celebrate India's rich linguistic and cultural heritage. It serves as a multilingual Corpus Collection Engine, enabling users to contribute and explore proverbs, slang words, and tongue twisters across 10+ Indian languages and dialects.

🚀 Features 🌐 Multilingual Support – Hindi, Telugu, Tamil, Bengali, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Odia, and more

🤖 AI-Powered Intelligence – Mistral AI (primary), OpenAI (fallback), Ollama (local)

🗳️ Community Contributions – Submit, vote, and categorize cultural expressions

📊 Admin Dashboard – User stats, language analytics, CSV/JSON export

🤗 Hugging Face Integration – Dataset sync, backup, and model-ready formats

📱 Modern UX – Mobile-friendly, offline-first, PWA, dark/light modes

🔐 Data Governance – Validation, version control, and privacy safeguards

ManaBhasha AI – Telugu Q&A Chatbot ManaBhasha AI is a question-answering chatbot built for Telugu speakers. It leverages Retrieval Augmented Generation (RAG) architecture with IndicBERT embeddings to deliver accurate, context-aware answers in Telugu. This project is part of the Summer of AI 2025 Hackathon and aims to contribute to building open-source language models and tools for regional language AI.

OCU-AI is a smart healthcare assistant that combines Natural Language Processing (NLP) and Computer Vision to assist users in understanding medical conditions and detecting eye diseases from retinal images.

This web-based application has two core components:

Medical Q&A Chatbot

Uses Retrieval-Augmented Generation (RAG) with LangChain, Pinecone, and Phi-3 LLM.

Answers user queries using embedded medical knowledge extracted from PDF files (like textbooks or research).

Designed to explain medical terms in simple, non-technical language.

Always includes a disclaimer to consult a qualified medical professional.

Eye Disease Detection Model

Allows users to upload retina images.

Predicts diseases like:

Cataract

Glaucoma

Diabetic Retinopathy

Normal (Healthy)

Uses a Keras CNN model trained on retina datasets.

After prediction, the chatbot explains the condition using natural language.

This project bridges AI-powered document retrieval, medical imaging, and LLM-based explanation, offering a foundation for real-world smart health applications.

The Meme Generator is an interactive web application that leverages the power of AI to generate witty captions and customize memes. Using OpenAI's GPT-3 model, the app creates engaging, humorous captions for a variety of popular meme templates or user-uploaded images. The app also features multilingual support, enabling users to translate captions into multiple Indian languages, making it ideal for a diverse, global audience.

📘 Syllabus Sakhi AI – Open-Source AI Assistant for Competitive Exams Syllabus Sakhi AI is an open-source bilingual assistant designed to help competitive exam aspirants—especially Telugu-speaking learners—explore and understand the complete syllabus of various Central and State government exams in India.

🎯 Purpose & Target Audience The assistant was created with the needs of rural and regional-language learners in mind, especially those:

Preparing for UPSC, SSC, RRB, APPSC, TSPSC, TET, DSC and other government exams.

Struggling with English-dominated resources.

Seeking structured, topic-wise syllabus coverage in Telugu or English.

Supported exams include:

UPSC Civil Services (CSE)

SSC CGL, CHSL, MTS, CPO

RRB NTPC, Group D

AP TET, DSC, APPSC Groups 1–4

TS TET, DSC, TSPSC Groups 1–4

🔑 Key Features 🗣️ Bilingual Support: Telugu (default), English, or mixed bilingual responses.

📚 Comprehensive Coverage: 19 exams with Prelims/Mains + section-level syllabus.

📄 Structured Knowledge Base: Markdown files with detailed topic lists & weightage.

🔍 Hybrid Search: Combines keyword + semantic retrieval for improved accuracy.

🌐 Language Selector: UI lets users pick TELUGU or ENGLISH on startup.

⚠️ Graceful Fallbacks: Handles missing syllabus gracefully with polite error responses.

This project aims to make syllabus navigation simple, accessible, and local-language friendly, empowering more students to prepare confidently for government jobs.

SciBot is an intelligent, RAG-based research assistant designed to retrieve and answer questions from PDF documents with high accuracy. This system combines the power of LLMs (Large Language Models) with a vector database to perform context-aware question answering directly from PDF data, making it ideal for students, researchers, and professionals.

🧠 Key Features: 📄 PDF Ingestion & Parsing: Automatically reads and extracts structured text content from uploaded PDF files using tools like PyMuPDF or pdfplumber.

🧹 Text Preprocessing: Cleans the extracted text, chunks it into semantically meaningful passages, and removes irrelevant formatting or metadata.

🔍 Embedding Generation: Converts text chunks into numerical vector embeddings using OpenAI Embeddings API, HuggingFace models, or Instructor embeddings.

🗂️ Vector Database Storage: Stores these embeddings in a fast vector database like ChromaDB, FAISS, or Pinecone, enabling efficient semantic retrieval.

🤖 RAG-based Question Answering:

When a user asks a question, relevant chunks are retrieved from the vector DB.

These are passed along with the query to an LLM (like GPT-4, OpenAI GPT-3.5, or Gemini).

The model then generates accurate, grounded responses using the retrieved context.

🧪 Research-Oriented UI:

Built with Streamlit for a fast, minimal interface.

Users can upload PDFs, type questions, and get natural language answers.

Highlighted references or citations from the PDF