Back
EN
SW
EN
SW

Hopsworks AI Lakehouse embeds NVIDIA NIM for LLMs, helping customers boost generative AI initiatives

July 1, 2024

Hopsworks is embedding NVIDIA NIM inference microservices into its leading AI Lakehouse. Tasks such as finding requirements out of legal and compliance documents, finding check or claims fraud, screening negative news, and creating agile development, user stories and documentation once took armies of knowledge workers months to implement. Now they take days, thanks to the Hopsworks AI Lakehouse.

With the Hopsworks AI Lakehouse, users can see more accurate and reliable generative AI by leveraging all of the capabilities of the platform, including the function-calling capabilities of NVIDIA NIM, Hopsworks’ high-performance data layer, and Hopsworks’ fully integrated platform services. Also incorporated into the platform are NIM support on open-source KServe, observability metrics, and support for PEFT models.     

Hopsworks unlocks the full potential of function calling by enhancing retrievals with context from structured data such as user behavior and context, recent documents not included in the LLM, and frequently asked questions and answers. The Hopsworks AI Lakehouse is also fully customizable and can be quickly extended to any number of related use cases that leverage the structured data and embeddings that are managed by the Lakehouse. Hopsworks AI Lakehouse is the most trusted platform available, providing enterprise-grade security and governance, zero downtime, and the capability to run across the cloud and on-premise data centers. The Hopsworks AI Lakehouse is a platform for quickly developing, deploying, and managing any ML or GenAI driven application, and it also delivers pre-built application accelerators for:

  1. Legal & Compliance
  2. Agile Development Automation
  3. Check Fraud
  4. Claims Fraud
  5. Healthcare Imaging using NVIDIA MONAI Toolkit
  6. Negative News Screening

NVIDIA NIM, part of the NVIDIA AI Enterprise software platform, is a set of easy-to-use microservices that accelerate the deployment of generative AI models across the cloud, data center, and work stations. NIMs are the accelerated computing engines of the new intelligence economy, connecting the power of the latest AI foundation models — securely deployed on NVIDIA accelerated infrastructure — with enterprise customers everywhere. 

Benefits

  • Maintain security and control of generative AI applications and data with self-hosted deployment of the latest AI models in a user’s choice of infrastructure, on premises or in the cloud 
  • Speed time to value with prebuilt, cloud-native microservices that are continuously maintained to deliver optimized inference on NVIDIA accelerated infrastructure 
  • Empower enterprise developers with industry-standard APIs and tools tailored for enterprise environments 
  • Improve TCO (Total Cost of Ownership) with low-latency, high-throughput AI inference that scales with users’ cloud infrastructure
  • Leverage enterprise-grade software with dedicated feature branches, rigorous validation processes, and support, including access to NVIDIA AI experts and defined service-level agreements 

 Description of Integration 

  • Hopsworks’ integration of NIM into the Hopsworks AI Lakehouse offerings, described above as part of Hopsworks’ managed service, is available either on premises installed in a customer data center, or inside a customers’ private or public cloud. The Hopsworks AI Lakehouse is a managed operational AI platform for the development and deployment of ML predictions or generative AI products from raw data sources to the applications and services. Integrated platform services include a data layer (online and offline store, vector index, knowledge graph), compute, NIMs and other models, API, collaborative UI, version control, data validation, orchestration, search and discovery, and security & governance.
  • The Hopsworks AI Lakehouse provides new capabilities by going well beyond simple storage and vector indexing to also coordinate function calling of structured context in relational databases as well as knowledge graphs. Hopsworks can demonstrate the dramatic improvements from function calling in the Compliance, Check Fraud, Claims Fraud, Agile Development Automation, and Negative News Screening demos.

 How to get started

  • Get started with NVIDIA NIM at ai.nvidia.com and sign up here for a demonstration of the Hopsworks Generative AI Lakehouse.
No items found.