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.

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

ripgrep for your team's knowledge base — hybrid retrieval over Slack, Confluence, Drive, Git, and files. Self-hosted, single command.

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

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.

Self-hosted AI evaluation framework — run prompts across Claude, GPT-4o, and Gemini in parallel with 4 scoring methods, cost/latency tracking, and live SSE streaming.

Citation-grounded, refusal-bounded retrieval for regulated-domain corpora. Citation grounding and refusal-on-low-confidence enforced as deterministic checks in code, not prompt instructions. v0.1: FDCPA.

RAG-powered prior-art patent search assistant — vector retrieval + Cohere rerank + Claude streaming, deployed on AWS via CDK

MongoDB Atlas + Voyage AI demo: hybrid vector/lexical search and template-driven rationale for healthcare prior authorization

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

Agentic RAG with clickable citations and synced document viewer

Semantic talent search & ranking API: FastAPI, ChromaDB, Voyage embeddings, optional Anthropic rerank.

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.

Persistent per-project memory for Claude Code via MCP server (4 tools) + 4-layer L0/L1/L2/L3 extraction pipeline. English-localized coder-focused fork of Tencent/TencentDB-Agent-Memory with vector recall, Voyage embeddings, and SessionStart auto-injection.

See how chunking strategy changes RAG retrieval. Same document, same question, different chunks → different answer. Built with Next.js, Voyage embeddings, and Claude.