swirlai/swirl-search

AI Search & RAG Without Moving Your Data. Get instant answers from your company's knowledge across 100+ apps while keeping data secure. Deploy in minutes, not months.

What it solves

SWIRL provides federated AI search and Retrieval-Augmented Generation (RAG) across multiple applications and data sources without requiring users to move, duplicate, or copy their data into a central vector database. It eliminates the need for complex ETL pipelines and allows search results to respect the original source's existing permissions.

How it works

Instead of indexing data beforehand, SWIRL queries sources live. It uses a pipelined Processor architecture to transform queries and responses. It then re-ranks results using cosine vector similarity (via spaCy and NLTK) and can optionally generate an answer with citations using a connected LLM (such as OpenAI).

Who it’s for

Teams and developers who need a unified search interface across tools like SharePoint, Confluence, GitHub, and Jira, and those who want to implement RAG without the infrastructure overhead of managing a vector database.

Highlights

  • Zero Data Movement: Data stays in place, avoiding the second copy to govern and secure.
  • Federated Search: Access to over 100 connectors for various apps and services.
  • Real-time RAG: Generates grounded answers with citations using your own LLM key.
  • Extensible Architecture: Easily extensible Connector and Mixer objects for adding new data sources.
  • Quick Deployment: Can be deployed via Docker in approximately two minutes.

Related

  • Project
    pathwaycom/llm-app

    A collection of ready-to-deploy AI pipeline templates for high-accuracy, real-time RAG and enterprise search that syncs directly with live data sources.

  • Project
    neuml/rag

    A Streamlit application backed by txtai that implements both Vector and Graph RAG to provide factually grounded LLM responses.

  • Project
    infiniflow/infinity

    An AI-native database designed for LLM applications that provides high-speed hybrid search across dense and sparse embeddings, tensors, and full-text data.

  • Project
    HKUDS/LightRAG

    A lightweight, graph-based RAG framework that combines knowledge graphs and vector embeddings to provide efficient, high-quality retrieval and global context understanding.

  • Project
    Bessouat40/RAGLight

    A lightweight and modular Python library for implementing Retrieval-Augmented Generation (RAG), allowing users to easily integrate various LLMs and vector stores to chat with their documents.