Pipeshub

Pipeshub is an open-source, self-hosted web app and server that serves as an AI-native execution layer for enterprise knowledge. It connects fragmented business data across various applications to provide a unified context layer for search, question and answering, and deep research. By aggregating data from multiple sources, it allows teams to interact with their internal documentation and communication history through a single interface.

The software is deployed as a server application, typically using Docker Compose for local or VPC installations. It allows organizations to maintain full control over their data by supporting a bring-your-own-model approach, ensuring that sensitive information remains within the internal infrastructure. The deployment process includes an interactive installer that manages secrets, graph database configuration, and image tag selection.

Key features

• Grounded answers with precise block citations to original documents
• Permission-aware search that enforces source-level access controls
• Graph-backed retrieval to capture relationships across enterprise data
• Over 30 enterprise connectors with real-time and scheduled indexing
• No-code visual builder for creating AI agents and actions
• Safe execution sandbox for generating reports, charts, and dashboards
• Multimodal understanding of images, diagrams, and scanned files
• Support for various file formats including PDF, Office documents, and Google Workspace

The architecture utilizes a Python-based backend with FastAPI, LangChain, and LangGraph. It employs Qdrant for vector similarity search and Neo4j or ArangoDB for graph database capabilities. For data ingestion and processing, it uses Docling and pdfplumber to extract content from diverse document types. The system is designed for developers and engineers, offering APIs, SDKs in Python, TypeScript, and Go, and support for the Model Context Protocol (MCP) to integrate enterprise context into other AI workflows.

Pipeshub is an extensible workplace AI platform for teams requiring a private, explainable knowledge management system that avoids the data privacy risks of third-party cloud AI services.