A health conversational assistant built on LangGraph and multi-source RAG: graph-orchestrated intent routing; retrieval fused with RRF across a Neo4j knowledge graph, Chroma vectors, and keyword search; Redis-backed memory and user profiling; supports subgraphs such as diet, exercise, multimodal, nearby maps, and user profile; decoupled FastAPI + R

Multimodal RAG AI Assistant for knowledge-based question answering using vector search, retrieval pipelines, and LLM reasoning.

No description available

No description available

No description available

No description available

A sophisticated Hybrid RAG System & AI Assistant. It prioritizes a local knowledge base (PDF/OCR) for secure, document-specific retrieval, falling back to Gemini API only when necessary. Features multimodal OCR for handwritten/scanned analysis, automated SMTP email integration, and a high-performance vector-search architecture. 🧠📄🚀

A multimodal Retrieval-Augmented Generation (RAG) assistant that lets you upload, embed, and query your own knowledge base.

A multimodal RAG AI assistant that extracts knowledge from PDFs and images and answers questions using OpenAI, LangChain, and ChromaDB.

JAN (Just Another Neuralnet) - An intelligent multimodal AI screen overlay assistant built with React 19 and Gemini Multimodal Live API. Features real-time screen perception, voice/text interaction, image/video generation, and RAG-powered knowledge base.

This "Context-Aware Intelligent Knowledge Assistant (Multimodal RAG)" is an excellent, interview-winning, production-ready full-stack + AI project. It directly solves enterprise hallucinations by grounding answers in user-uploaded PDFs/docs + YouTube videos.

QueryBox is a multimodal AI learning assistant for technical learners that ingests handwritten notes, images, and PDFs, converts them into structured knowledge using OCR and RAG, and enables context-aware learning through an LLM-powered chat interface.

A Retrieval-Augmented Generation (RAG) knowledge assistant that leverages multimodal data sources and AI to answer questions intelligently. The project includes integration with OpenAI APIs (API keys kept secure locally) and demonstrates secure handling of environment variables.

A multimodal AI research assistant that combines language models, vision-language models, and diffusion models to process text, images, and audio. The assistant utilizes agentic AI frameworks for task-specific agents, integrates RAG for up-to-date knowledge retrieval, and maintains contextual memory for personalized interactions.

An AI-powered RAG chatbot that combines document search with YouTube video transcription using Hugging Face Whisper. Users can upload PDFs, text files, and add video links for seamless, accurate Q&A. Built with LangChain, ChromaDB, and Ollama LLM, it offers a versatile, multimodal knowledge assistant.