ServiceNow today announced its latest generative AI solution, Now Assist for Virtual Agent, with the aim of revolutionizing self-service by offering intelligent and relevant conversational experiences. The new capability expands on ServiceNow’s strategy of integrating generative AI capabilities into its Now Platform, which helps customers to streamline digital workflows and optimize productivity.
This tool utilizes generative AI to deliver direct and contextually accurate responses to user inquiries. Integrated with the Now Platform, it will enable users to swiftly access relevant information and connect with digital workflows tailored to their needs. Now Assist provides user assistance with internal code snippets, product images or videos, document links and summaries of knowledge base articles.
According to the company, this self-service capability will help users obtain quick and accurate solutions, even when they need guidance on whom to approach or where to begin. The company believes that by enhancing self-solve rates and accelerating issue resolution, the feature significantly boosts productivity.
“One of the key goals of our new offering is to unlock additional productivity without added complexity by providing direct, relevant conversational responses,” Jeremy Barnes, VP for platform product AI at ServiceNow, told VentureBeat. “By connecting exchanges to automated workflows, customers can get the information they need within the context of their organization.”
ServiceNow’s launch of Now Assist aligns with the introduction of their Generative AI Controller, which serves as the foundation for all generative AI functionality on the Now Platform. In addition, the company has also collaborated with Nvidia to develop customized large language models (LLMs) for workflow automation.
Leveraging generative AI to streamline user inquiries
Now Assist for Virtual Agent can be easily configured using Virtual Agent Designer in a low-code, drag-and-drop environment. Additionally, users can create and deploy conversational self-service with the tool’s diagram drag-and-drop designer, which incorporates natural language understanding (NLU).
ServiceNow says this integration can be easily incorporated into an organization so they can begin automating and streamlining digital workflows to achieve faster responses.
“Now Assist allows organizations to easily connect across a company’s internal knowledge base, and then supplement answers with general purpose LLMs like Microsoft Azure OpenAI Service LLM and OpenAI API,” said Barnes.
In partnership with Nvidia, the company is actively developing custom LLMs tailored specifically for ServiceNow. These LLMs will be readily available and integrated into the Now Platform.
Barnes highlighted that the company’s strategy encompasses supporting both general-purpose LLMs and providing domain-specific LLMs. The ongoing collaboration with Nvidia aims to address a broad spectrum of customer requirements with custom LLMs.
Custom LLMs built with Nvidia
The company is developing custom LLMs using Nvidia’s software, services and infrastructure, trained on data specifically for the ServiceNow Platform, Barnes explained.
“We believe there will be many more exciting advances as we continue to strengthen workflow automation and increase productivity,” he said.
Barnes explained that if an organization’s knowledge base lacks sufficient information to provide a contextual response to a general question, Now Assist will establish a connection with general-purpose LLMs to augment the answer.
“If a user doesn’t know who to ask or where to start, our new solution will help them quickly determine the most relevant answer without having to scroll through endless links or knowledge base articles,” Barnes added. “For our customers, this is about simplification and not having to slow down to understand how and where to get the help you need — but to be able to get it at the speed of your work.”
The company said that Now Assist for Virtual Agent and Now Assist for Search are presently accessible to a select group of customers and are anticipated to be widely available in ServiceNow’s Vancouver release scheduled for September 2023.
What’s next for ServiceNow?
Barnes said that ServiceNow is actively exploring future use cases of generative AI to enhance productivity across various business functions, such as IT, employee experience and customer service.
“We are exploring additional future use cases to help agents more quickly resolve a broad range of user questions and support requests with purpose-built AI chatbots that use LLMs and focus on defined IT tasks,” he said. “Internally, ServiceNow is exploring how AI can be used to generate and document code and scripts as well as evaluating how it can help employees find information faster for things like benefits, PTO policies, opening incidents and more.”
The company aims to integrate all workflows with generative AI and low code. By doing so, ServiceNow believes it will unlock new use cases that effectively leverage the technology’s potential across industries and enable the creation of new revenue streams.
“We’re incredibly excited about enterprise AI,” said Barnes. “There are hundreds of use cases where generative AI — applied to a business problem you’re solving for — can radically transform the productivity curve.”
