Chipper offers a modular, containerized web interface and CLI for building and deploying RAG (Retrieval-Augmented Generation) pipelines, document processing, and web scraping. It targets tinkerers, educators, and developers seeking to enhance local or cloud-based LLMs with custom knowledge bases, providing an extensible platform for AI exploration and integration with tools like Ollama and Haystack.

How It Works

Chipper leverages Haystack for RAG pipeline orchestration, Ollama for local LLM inference, and Elasticsearch for vector storage. It functions as a proxy for the Ollama API, enabling third-party clients to access RAG capabilities. The architecture supports document chunking, web scraping, and audio transcription, with a focus on offline functionality via a vanilla JavaScript/TailwindCSS web UI and Edge TTS for client-side speech synthesis.

Quick Start & Requirements

• Install/Run: docker compose up
• Prerequisites: Docker, Docker Compose. Ollama instance recommended for local LLM use.
• Links: Project Site, Live Demo

Highlighted Details

• Ollama API proxy with Bearer token authentication.
• Offline-capable, framework-agnostic web UI.
• Supports local (Ollama) and remote (Hugging Face) model integration.
• Distributed processing via chained Chipper instances.

Maintenance & Community

• Active development by a single primary contributor.
• Integrates with several community projects: page-assist, open-webui, Enchanted, Ollamac.
• Roadmap includes expanded client settings, tutorials, and automated testing.

Licensing & Compatibility

• MIT License. Permissive for commercial use and closed-source linking.

Limitations & Caveats

The project is explicitly noted as a personal project not designed for commercial or production use, recommending users conduct their own due diligence for such scenarios. A React-based web application is listed as an upcoming feature, suggesting the current UI may be less polished or feature-rich than a modern SPA.