NVIDIA has announced the development of RAG-based question-and-answer LLM workflows, aimed at enhancing AI capabilities and user experiences. The initiative leverages Perplexity's search API to facilitate web search and summarization tasks while minimizing latency and token usage. NVIDIA's approach utilizes NIM microservices and A100-equipped nodes for efficient model deployment.

Source: https://Blockchain.News/news/nvidia-develops-rag-based-llm-workflows-enhanced-ai-solutions