NVIDIA introduces NeMo Retriever NIM microservices for enhanced AI applications
From NVIDIA: 2024-07-23 11:15:16
NVIDIA has introduced new NeMo Retriever NIM microservices to enhance accuracy and scalability for AI applications by enabling retrieval-augmented generation. With NeMo Retriever, developers can connect custom models to diverse business data, boosting accuracy for applications like customer service chatbots and security vulnerability analysis. These microservices are available now.
NeMo Retriever NIM microservices include embedding and reranking models that offer high accuracy for text question-answering retrieval. The microservices are transparent, reliable, and provide state-of-the-art models for developers creating AI applications. By combining embedding and reranking models, NeMo Retriever ensures the most helpful and accurate results for enterprise usage, outperforming other models in accuracy.
The NeMo Retriever microservices power a variety of AI applications such as chatbots, data analytics, and retail shopping advisors. Integration of NeMo Retriever into platforms like DataStax, Cohesity, Kinetica, and NetApp helps companies develop generative AI applications that provide accurate insights for their customers and users. Global system integrator partners are also working to incorporate NeMo Retriever NIM microservices for their enterprise clients.
NVIDIA’s NeMo Retriever NIM microservices can be utilized alongside Riva NIM microservices to enhance speech AI applications like text-to-speech and automatic speech recognition. The microservices offer flexibility for developers to integrate community models, NVIDIA models, or custom models in cloud, on-premises, or hybrid environments, ensuring a modular approach to building AI applications. Enterprises can deploy NIM microservices through the NVIDIA AI Enterprise software platform on various accelerated infrastructure options.
Read more at NVIDIA: New NVIDIA NeMo Retriever Microservices Boost LLM Accuracy and Throughput