ServiceNow's Vancouver release introduces transformative generative AI capabilities. To unravel its potential, INRY is rolling out articles illuminating its features, debunking myths, and presenting real-world applications.
Notably, AI isn't new to ServiceNow. Over the past six years, they've integrated several AI-driven technology firms, each bringing impactful workflow automation tools. This purpose-built AI, informed by specific use cases, ensures maximum impact.
AI was a hot topic at Knowledge this year. The first announcement made by ServiceNow presented two new solutions: ServiceNow Generative AI Controller and Now Assist for Search. ServiceNow quickly followed with an announcement of a new partnership with Nvidia.
ServiceNow announced a third solution in its generative AI capability set: Now Assist for Virtual Agent.
The fourth announcement that we'll feature is the release of the StarCoder Large Language Model. For the upcoming articles, we will dive into the details of these announcements and other AI use cases on the ServiceNow platform.
ServiceNow will make use of high-performance software services and infrastructure from Nvidia for the development of custom LLMs. These LLMs will be trained on data that come specifically from the workflows of the Now platform. The outcome of this will be expanded AI functionality for all enterprise segments in which ServiceNow provides workflow solutions. "Bring your own LLM" – We think domain-specific LLMs will provide big value for our customers going forward.
The ServiceNow Generative AI Controller is a major new addition to the Now platform. This controller will allow your organization's instance to connect to large language models (LLMs) via API. The two services highlighted in the announcement are OpenAI (AKA Chat GPT) and Microsoft's OpenAI. The controller will also enable connection to domain-specific models. This is where the work being done with Nvidia comes into consideration for platform-specific workflow data that needs to be kept separate from the public general-purpose models. Also, here is where the StarCoder LLM for text-to-code comes into play.
The Now Assist for Search application will leverage the LLM connections of the controller to provide natural language responses to user inquiries on the Now platform. These can come from the search interface of the platform, service portals, the search interfaces of the Next Experience UI, or the platform's virtual agent chatbot interface. Because the search functionality here pulls information from the data of your organization's own instance, the results can be more accurate and directly relevant to your user inquiries rather than relying solely on the information of general-purpose models.
The Now Assist for Virtual Agent will leverage the prior two solutions, the controller and the search capabilities, to provide natural language response in the virtual agent chatbot. Because a Virtual Agent is a programmable intelligence component, the points where generative AI is placed into a virtual agent conversation flow are fully designable. This will be done using the same Virtual Agent Designer application that platform administrators, developers, and process owners are familiar with today.
The last announcement was the StarCoder LLM for code generation, released by Big Code. Big Code is an open scientific collaboration led by ServiceNow Research and Hugging Face, working on the responsible development of code LLMs and empowering machine learning and open-source communities through open governance.
Here is a preview of the upcoming articles to dive into these new capabilities and guide you as you establish your organizational practices.
- Up and running with new Generative AI capabilities in ServiceNow
- Accelerate productivity with Generative AI on the Now Platform
- Deep dive into the Generative AI Controller for ServiceNow
- Deflect cases and empower people with Conversational Exchanges with Now Assist
- Generate content automatically leveraging AI on the Now Platform
- ServiceNow AI Search - Reduce effort and maximize productivity
- Avoid Human Errors with Document Intelligence
- Boost developer productivity with Text to Code
- Reduce manual work for agents with Case Summarization
- Task Intelligence Releases New Automation Tools for ITSM
We look forward to exploring these new opportunities and helping you navigate your AI Journey!