A few weeks ago, we announced the newest offer in our Dell APEX Cloud Platform family, APEX Cloud Platform for Red Hat OpenShift, jointly engineered with Red Hat to transform OpenShift on-premises. Today we’re excited to share the publication of the Dell Validated Design for Red Hat OpenShift AI on APEX Cloud Platform, which shows how you can build a digital assistant customized with your business’s data.
There is incredible enthusiasm around the possibilities of artificial intelligence (AI), with 75% of organizations increasing investments in AI initiatives.* While AI offers enormous potential, organizations are being challenged to find ways to deliver value with AI for their business. Some early successes with digital assistants include improving clinical trials for drug development, enhancing customer service in retail and simplifying access to government services.
Unlocking the Business Value of GenAI
It’s because of the value delivered across a wide array of industries that digital assistants are one of the most popular use cases for generative AI (GenAI). The technology is designed to assist users by answering questions and processing routine tasks. In their simplest form, you might think of a digital assistant as a chatbot. And while chatbots have been used for many years, they have had significant limitations (we’ve all had at least one less-than-ideal customer service experience with a chatbot). GenAI revolutionizes how this type of technology works and what it’s capable of doing in several ways.
First, there are large language models (LLM), trained on massive amounts of data to understand existing content and generate new content. LLMs are remarkable and generate incredible answers, which by themselves can lack domain-specific information for your business and do not stay up-to-date on their own. That’s where retrieval-augmented generation (RAG) comes in. With RAG, you can augment the LLM with your company’s information, rapidly training it with relevant information that stays current.
Making GenAI Accessible
In our validated design we showcase this framework by augmenting Llama 2, a popular LLM, with a series of Dell documents. By doing so, users get more accurate information to their queries and can see where the information came from with direct access to the source document. The video clip below shows the impact of RAG and how the response to the same question improves once a document has been ingested.
While this is exciting, we know getting started with GenAI projects can seem daunting. That’s why our validated design provides step-by-step instructions for how a data scientist could take advantage of an LLM with RAG to build their own digital assistant using Red Hat OpenShift AI on the APEX Cloud Platform. We utilized a series of open-source operators to make it easy to replicate and tweak the design to fit your specific business’s needs. Red Hat OpenShift AI represents an alternative to prescriptive AI/ML suites, providing a set of collaborative open-source tools and a platform for building models without worrying about the infrastructure or lock-in from public cloud-specific tools. By using Red Hat OpenShift AI on APEX Cloud Platform, you can get started quickly, focusing on how to deliver valuable business outcomes rather than how to stand up and maintain a complicated series of individual tools.
Our Services team is also well-versed in unleashing the value of your data with context-specific GenAI, like RAG, and can help guide your organization from strategy and implementation to adoption and scaling of GenAI solutions.
Getting Started with a Digital Assistant
There is a lot of excitement around GenAI. However, getting started can be challenging. The Dell Validated Design for Red Hat OpenShift AI on APEX Cloud Platform offers organizations a simple and straightforward way to start getting value with GenAI by implementing a digital assistant. Click here to see the full validated design and here to learn more about APEX Cloud Platform for Red Hat OpenShift.
*Dell Technologies Generative AI Pulse Survey, August and September 2023, www.dell.com/GenAIPulse