Code search MCP for Claude Code. Make entire codebase the context for any coding agent.

An interactive RAG agent built with LangChain and MongoDB Atlas. Manage your knowledge base, switch embedding models, and tune retrieval parameters on-the-fly through a conversational interface.

Polymarket, bulit for Move, powered by AI

A retriever pipeline Python web application that allows users to chat with PDFs and web articles using AI. Combines FAISS vector similarity search and BM25 keyword matching for accurate document retrieval and uses Voyage AI embeddings and LLaMA hosted on Cloudflare Workers AI for question answering.

An autonomous, state-persistent recovery agent for stranded travelers. Powered by MongoDB Atlas Vector Search & LangGraph.

This demo showcases a store associate application built on MongoDB Atlas, created to streamline product discovery and inventory visibility as part of a unified commerce strategy.

This demo showcases how retailers can turn digital receipts into personalized, AI-powered experiences using MongoDB Atlas. Built with event-driven microservices, this solution demonstrates how to centralize invoice data and use it in real time to enrich customer journeys.

Local document indexer MCP server for semantic search over PDF, Excel, SQL, Markdown, and HTML files using Qdrant and Voyage AI embeddings.

Build and manage intelligent autonomous agents using a modular, multi-language framework in Python and Rust.

A Technical RAG-DEMO

Trying to understand RAG

Multi-path RAG agent designed to provide support to dementia caregivers.

Real-Time Lakehouse on Azure & Databricks. Ingests clickstream events via Event Hubs & Delta Live Tables (DLT) with <500ms latency to trigger Agentic AI recovery workflows.

3-stage candidate search pipeline: Voyage-3 vector retrieval, hard-criteria filtering, and GPT-4o-mini reranking across 10 role configurations

Self-hosted code intelligence hub. Uses VoyageAI embeddings for semantic searching code. CLI + server modes, MCP for Claude Desktop, FTS/regex, git history search, and SCIP searches. Automatic inter repository domain mapping. Automatic wiki generation for repos containing .md files and images.

Use AI and real-time data processing to prevent customer churn in e-commerce. Monitor customer behavior through MongoDB change streams, detecting critical behavior signals. When these moments are identified, an AI agent generates personalized NBAs.

Production-grade Neo4j GraphRAG pipeline — async scraping, LLM entity extraction with gleaning, Leiden community detection, Voyage AI embeddings, 10-step entity normalization, and comprehensive validation

Stream, embed and store your S3 buckets onto Mongo Atlas 🍃

🧠 Κορυφαίο RAG σύστημα για Obsidian • Συνεργασία Ανθρώπου-AI • Late Chunking • HyDE • Self-Correction • Graph Retrieval

RAG-powered codebase explorer — index any GitHub repo and ask natural language questions with cited file paths and line numbers.