The Litmos AI Assistant is designed to generate responses strictly from the content available within your organization’s Litmos LMS environment. It does not search the web or use external data sources; all generative responses are based solely on the courses, training materials, and documents that have been uploaded and made available within your Litmos instance. Each response is accompanied by a course citation, ensuring that the source of information is always known and traceable. Litmos has integrated advanced large language models (LLMs), specifically Llama 3.1, to power its AI assistant. This model was selected after comparative testing for its superior instruction-following, reliability, and accuracy over previous models and competitors. The AI assistant uses retrieval-augmented generation (RAG), which means it retrieves relevant information from your internal content and then generates responses, further enhancing accuracy and contextual relevance.