A travel assistant made to answer travel related queries. It responds using data retrieved from both Pinecone and Neo4j. Currently only works for Vietnam

Q&A inteligente sobre o Livro Amarelo — explore propostas e ideias para o futuro do Brasil.

EduPilot is an intelligent course assistant built for Indiana University students. It answers questions across four graduate courses — AML, ADT, STAT, and LLM — using a seven-stage multi-agent RAG pipeline that retrieves directly from course lecture slides and materials, cites every claim and verifies its own answers, and refuses to hallucinate.

This is a production deployed RAG system that answers medical questions grounded in verified clinical documents — not hallucinations.

smolchat is a full-stack agentic AI platform (LangChain + Next.js) with multi-LLM chat, long-term cross chat memory, STT & TTS, RAG (multi-format doc support) backed by Pinecone VectorDB, and client-side chat persistence via IndexedDB. The smolchat agent integrates with tools like YouTube search, Web Search, Web Page Search, and Wolfram Alpha KE.

An AI-powered web application that lets users upload PDF documents and have intelligent conversations with them. Ask questions and get instant accurate answers with page references.

AI-powered meeting intelligence platform with transcription, RAG-based chat, action item extraction, Slack integration, and semantic search using OpenAI, Pinecone, FastAPI, and Next.js.

A Retrieval-Augmented Generation (RAG) pipeline built to query and analyze state-level energy policies.

Built a multi-agent AI Assistant using LLMs, LangChain, and LangGraph with RAG-based document retrieval. Enabled context-aware Q&A over PDFs using Pinecone vector DB. Implemented FastAPI backend with JWT authentication for scalable, secure AI interactions.

An enterprise-grade code intelligence platform that automates PR reviews, visualizes architectural dependencies, and actively generates self-healing patches using advanced AI engines and context-aware RAG.

Smart Photo Organiser is a full-stack face recognition web app that identifies people in group photos. Register faces with a name and photo — then upload any group image and the app detects every face, matches each one against the database using vector similarity search, and returns the top 3 closest matches with confidence scores and a image

RAG pipeline to query and analyze software engineering case studies

Next.js v16, AI Applications with Gemni (Gemini 2.5) and OpenAI (GPT-5.0)

Built a RAG Pipeline ...

Multimodal RAG architecture integrating Pinecone vector store and Gemini LLM for querying text and images from documents.

In this project, I made a resume scoring system. And it not only scores candidates based on Job Description & Resume but it can also have user profiles and their personal hiring preferences. So a particular user can filter Resume in bulk according to their personal preference. It also learns user preferences with time and evolves for each user.

Four different types of Chatbots- one with a single pdf as context; one with a multiple docs of different types of docs as context; one with a enterprise knowledge; another with no pre context just like chatgpt and others;

AI credit-card recommender (RAG) with Mistral & AI credit-card recommender (RAG) with Mistral & Pinecone — semantic search, comparisons & FastAPI 🚀💳🔎

Provides the precautions, causes, remedies for the health upset given as input. Chatbot like interface.

Sec2ndBrain transforms your digital life into a chat-ready RAG knowledge base. Powered by Groq and Jina AI, you can instantly talk to your saved notes, YouTube videos, and Tweets.