Home > AI Solutions > Gen AI > Guides > Design Guide—Generative AI Digital Assistants in the Enterprise > What are digital assistants?
Our objective is to build an ethical and secure digital assistant interface platform that uses concepts originally developed for gamification applied to enterprise systems. By integrating Dell’s generative AI solution with an ecosystem of advanced AI software platforms, we can harness technology that was traditionally focused on gaming use cases and apply it to real-world business applications. The fully integrated solution designed by Dell Technologies brings digital assistants to the forefront of human interactions and drives operational efficiencies in both internal use cases and externally facing customer interactions.
By building this ecosystem and evolving the next generation of technologies, we aspire to bring the intelligence grid of the future to every customer and drive practical uses in retail, healthcare, and beyond.
The following figure depicts an example of a digital assistant that has been programmed to act as a technical support agent by supporting multiple modalities:
Depending on a user’s preference, the buttons at the bottom of the screen allow the system to turn on or turn off individual modalities.
Dell Technologies has partnered with UneeQ for our digital assistant frontend. UneeQ has differentiated itself from the competition by not only focusing on generative AI for life-like animations, but by also addressing the visual components associated with what people find to be unnatural. The nuances in the extent of lip movement along with emotion driven from the eyes, ears, and forehead are important. These nuances move people beyond the uneasy reaction that often occurs (also known as uncanny valley responses) at the first interaction. UneeQ is sensitive to cultural differences and has considered which animations are acceptable in one region compared to another.
The need for translation and speech in any language also requires that the synthetic generation dynamically adjusts not only to sentiment but also to movements that account for the user's background. As we focus on pushing the boundaries of human acceptance, Dell Technologies partners with UneeQ to bring our high-performance infrastructure and AI solutions expertise into the mix. This migration to an on-premises solution ultimately enables us to achieve response times in line with a truly interactive conversation: for each question asked, the latency of the answer is less than a few seconds.
For implementing a RAG approach, we have partnered with Pryon, which provides a commercial off-the-shelf AI-powered data retrieval product. The Pryon Retrieval Engine is a versatile AI tool that securely extracts answers from all types of content including audio, images, text, and video. Pryon ensures high-precision ingestion by emulating human-like document understanding, employing proprietary OCR technology to extract text in reading order from images, graphics, schematics, and handwritten notes. Pryon's retrieval process uses a proprietary combination of query expansions, out-of-domain detection, and query embedding to comprehensively understand natural language queries for ingested content matching. Then, Pryon uses three proprietary models to find and rank the best matched content quickly. All answers are presented with attribution to the source content.
Pryon seamlessly integrates with major enterprise knowledge base systems, enabling users to construct a unified retrieval knowledge base effortlessly without content consolidation. Pryon Retrieval Engine maintains document-level access control lists (ACL) and Pryon AI models do not train on enterprise data. This combination of features allows Pryon Retrieval Engine to offer industry-leading accuracy, at enterprise scale, with security controls, and fast time-to-value.