Newman An open source, local, AI assistant. RAG (Retrieval Augmented Generation) Uses the LaTeX papers stored in /data to augment the LLM response based on the prompt in retrieval.py. Papers can be downloaded for free from arXiv.