awslabs-lisa

A scalable infrastructure-as-code solution for self-hosting and orchestrating LLM inference with RAG capabilities, providing low-latency access to generative AI and embedding models across multiple providers.