Player not loading? Watch on YouTube
This tutorial, uploaded on August 27, 2024, demonstrates a local installation of Kotaemon, an open source interface for chatting with documents through retrieval augmented generation. No Kotaemon release number or separate recording date is supplied. The setup reflects the software shown at that time, rather than a current installation guide.
The presenter creates a Python 3.10 environment, clones the repository and installs its dependencies. He edits the environment configuration, adds PDF.js assets for the browser document viewer and starts the Python app. His demonstration uses an RTX A6000 GPU; the video does not establish that card as a minimum requirement. He also reports that some repository instructions did not work for him.
The document workflow covers uploading a PDF, waiting for indexing and asking questions about its contents. The interface displays the source document beside the answer. The presenter also shows reindexing, user settings and retrieval configuration, and describes his backend as GraphRAG.
For local LLM use, he selects Ollama in the settings and shows an existing Llama 3.1 installation alongside nomic-embed-text. Ollama installation itself is outside this walkthrough. OpenAI and Azure appear as alternative providers, so installing the interface locally does not by itself establish that every demonstrated configuration keeps processing on the same machine.
