Accenture and NVIDIA

There is no “easy” button when it comes to implementing generative AI solutions and strategies that drive business reinvention, but you can accelerate time to value and ease implementation efforts. Accenture is helping clients do just that by taking an active role as an integrator of the newly launched NVIDIA NIM Blueprints.

Getting started with Accenture and NVIDIA NIM Blueprints

NIM Blueprints are helping to redefine the landscape of enterprise AI development by bundling a set of ingestion patterns and ready-to-use, customizable AI workflows. Taking a step forward, Accenture is adding NVIDIA NIM Blueprints to Accenture AI RefineryTM, unveiled last month. Accenture AI Refinery enables clients to convert raw AI technologies and tools into scaled, enterprise AI systems. By creating the foundation for data-driven generative AI flywheels, these custom AI systems can continuously reinvent the highest-value enterprise capabilities.

Developers can use NIM Blueprints to accelerate their adoption of new technologies. These resources can help them reduce complexities of an oftentimes multistep process and allow them to modify existing assets using their data. NVIDIA’s catalog of pretrained, customizable AI workflows can empower developers by providing them with the tools needed to create and deploy advanced generative AI applications seamlessly across various industries, with a wide range of use cases.

The first batch of NIM Blueprints includes a digital human workflow that can be used to implement virtual agent services. It also includes a workflow to advance scientific discoveries in drug development through generative virtual screening, as well as a multimodal PDF data extraction workflow for enterprise retrieval-augmented generation (RAG). With RAG, companies can use generative AI applications to communicate with business data and synthesize more accurate responses.

NIM Blueprints are free for developers to experience, download and use. Developers can deploy them in production with the NVIDIA AI Enterprise software platform. We at Accenture specialize in guiding you on how to get started with this process.

More possibilities with NIM Blueprints and RAG

We know from our surveys and more than one thousand gen AI client projects that organizations across all industries are racing to find opportunities to create value through gen AI implementations. For example, many are pushing to enhance customer interactions by deploying humanlike AI agents. An NVIDIA NIM Blueprint powered by the NVIDIA Tokkio reference workflow supports implementation of 3D animated digital human interfaces that can help clients create more engaging user experiences. These are faster than generative AI-powered applications that offer just text or speech interactions.

The multimodal PDF data extraction for RAG NIM Blueprint uses NVIDIA NeMo Retriever microservices to unlock insights from massive volumes of enterprise data. With this enterprise-scale document-retrieval workflow, developers can create digital human technologies, AI agents or customer service chatbots that can quickly become experts on topics captured within their data universe. Enterprises can use the workflow to combine NeMo Retriever microservices with community or custom models to build high-accuracy, multimodal retrieval pipelines. These can be deployed wherever enterprise data resides.

Enterprises can also connect multiple NIM Blueprints to build increasingly sophisticated and specialized generative AI applications. For organizations aiming to integrate generative AI solutions and applications into their enterprise operations, this can offer:

• Customization and flexibility at the core: Each NIM Blueprint is equipped with sample applications, reference code, and detailed customization documentation. This ensures that developers have a solid foundation to start from as well as the flexibility to tailor the AI workflows to meet specific business requirements.
  
  
• Easy deployment across environments: With the help of complete deployment tools, the blueprints can be easily combined and integrated across different environments. Whether a client has an on-premises data center or uses a cloud platform, the NVIDIA AI Enterprise software platform can help ensure smooth deployment with optimal performance.
  
  
• Continuous improvement powered by continuous data collection: NIM Blueprints also help developers improve their applications throughout the AI lifecycle. As users interact with AI applications, new data is generated, which can then be used to refine and enhance the models in a continuous learning cycle. The blueprints help enterprises build their own generative AI flywheels with applications that link models with their data. NVIDIA NeMo facilitates this process, while NVIDIA AI Foundry serves as the production environment for running the flywheel.

Accenture is committed to helping our clients get started with adopting the latest technologies in generative AI. We’re working together with NVIDIA to provide more tools, resources and services to help users scale generative AI solutions and enhance customer experience.