Kotaemon: An Open-Source Tool to Chat with Your Documents 239

Kotaemon is an open-source tool that lets you converse with your documents. Built on RAG technology, it enables interactive dialogue with documents and supports multimodal retrieval, capable of handling different types of data such as text, charts, and tables.

Kotaemon Key Features:

• Host your own document QA (RAG) Web-UI. Supports multi-user login, organize your files in private/public collections, collaborate with others and share your favorite chats.

• Organize your LLMs and embedding models. Supports local LLMs and popular API providers (Openai, Azure, Ollama, Groq).

• Hybrid RAG pipeline. Sane default RAG pipeline with hybrid (full-text and vector) retrievers + re-ranking to ensure the best retrieval quality.

• Multi-modal QA support. Question answering over multiple documents with graph and table support. Support for multi-modal document parsing (optional option on the UI).

• Cite ahead with document preview. By default, the system provides detailed citations to ensure the correctness of LLM answers. View your citations (including relevance scores) directly in the in-browser PDF viewer with highlighting. Issue warnings when the retrieval pipeline returns low-relevance articles.

• Support for complex reasoning methods. Use question decomposition to answer complex/multi-hop questions. Support agent-based reasoning using ReAct, ReWOO, and other agents.

• Configurable Settings UI. You can adjust the most important aspects of the retrieval and generation process (including prompts) on the UI.

• Extensible. Built on Gradio, you are free to customize/add any UI elements as needed. Additionally, we aim to support multiple document indexing and retrieval strategies. A GraphRAG indexing pipeline is provided as an example.

Kotaemon Use Cases:

kotaemon serves as a functional RAG UI for end users who want to QA their documents and developers who want to build their own RAG pipelines.

For end users:

• A clean and minimalist UI for RAG-based QA.

• Supports LLM API providers (OpenAI, AzureOpenAI, Cohere, etc.) and local LLMs (via ollama and llama-cpp-python).

• Simple installation script.

For developers:

• A framework for building your own RAG-based document QA pipeline.

• Use the provided UI (built with Gradio) to customize and see how your RAG pipeline runs.

Kotaemon provides a clean and easy-to-use interface, supports the use of local LLMs or mainstream LLM API providers, and can be installed and used with a single click.

Kotaemon GitHub:https://github.com/Cinnamon/kotaemon