ServiceNow launched its Yokohama release of its Now Platform including AI agent orchestration, analytics and a workflow data fabric that integrates apps, data warehouses and lakes, and a workspace and studio to design and integrate agents.
The release highlights ServiceNow's broader strategy, which is to leverage its ability to tap into workflows and data across multiple enterprise systems and connect those dots to enable AI agents. ServiceNow, which scaled due to its role as a neutral workflow orchestrator, is betting it can also enable multi-agent flows across the Now Platform and third party apps and platforms.
ServiceNow’s Yokohama release is a fast follow up to its $2.85 billion acquisition of Moveworks.
"We are focused on a future where AI agents work autonomously with and for people to unlock outcomes and transform business," said Amit Zavery, ServiceNow's President, Chief Product Officer and Chief Operating Officer. "Right now, many AI agents are stuck in the same isolated systems that have created siloed ways of working for decades."
Simply put, ServiceNow is aiming to be the connective layer for agentic AI and adding context via the Now Platform's access to data, records and workflows. Yokohama will bake in AI agents throughout the ServiceNow platform to address use cases across IT, CRM, HR, security, finance and application development.
"We are building multi-agent systems to control workflows and independently solve tasks with proper governance," said Zavery. "Work doesn't really happen in silos, and neither does our AI, our orchestration capabilities connect every function through AI powered workflows."
ServiceNow's Yokohama release is also the first release under the company's hybrid pricing model that includes AI agents in Pro Plus and Enterprise Plus licensing models. As a result, enterprises will be able to use AI agents without additional charges, but there is a limit with consumption charges after that.
The release, which has ServiceNow's Workflow Data Fabric at its core, also has an updated Common Service Data Model (CDSM), a standardized framework for managing IT and business services.
CEO Bill McDermott has called ServiceNow's pricing model a win-win and a Goldilocks scenario in terms of enabling customers to adopt AI agents at their own pace without commitment up front. ServiceNow has more than a 1,000 customers using its AI agents.
Internally, ServiceNow said it has been using its AI agents to improve case deflection rates by 80% in the last 6 months in go-to-market operations, providing answers to seller questions 99% faster than request tickets and driving 20% productivity increases across HR and IT support.
Key parts of the Yokohama release include.
- AI Agent Orchestrator and AI Agent Studio are generally available for customers.
- AI Agent Orchestrator monitors agents, oversees them and develops plans for them to work together. ServiceNow's AI agents have the ability to connect to other systems. There are more than 50 integrations set for the AI agents in Yokoyama.
- AI Agent Studio enables customers to build agents with guardrails, workflows and actions and tools and data available.
- Service observability, which unifies multiple monitoring and observabilities tools in one dashboard.
- Preconfigured AI agents for security incident lifecycle, change management and proactive network test and repair. The autonomous change management AI agents generate custom implementation, test and backout plans based on impact, historical data and similar changes.
- Voice input for hands free interaction to summarize incidents and generate knowledge articles.
- AI agent analytics to gauge efficiency, productivity and alignment with enterprise KPIs.
- ServiceNow Studio, an AI-powered workspace where developers can use no-code, low-code and pro-code tools to build and deploy workflows and AI agents. ServiceNow Studio is integrated with AI Studio.
- The ability to create robotic process automation (RPA) bots with natural language to speed up development.
Early adoption
ServiceNow's Dorit Zilbershot, Group VP of AI Experiences and Innovation, said early adopters of AI agents are developing use cases with a "wide range of complexity."
Support is an obvious use case for AI agents, but one ServiceNow customer is using agents to navigate policies and how to address requests. That work is now automated, but it took hours for humans to address, said Zilbershot.
Sunil Tulyani, Service Management Platform Leader at Eaton, a power management company. Tulyani said Eaton started with ServiceNow for IT service management and then expanded into HR.
Eaton, which has been a ServiceNow customer, for more than two years has adopted Now Assist and the vendor's generative AI tools. "We gained the value from Now Assist already in this last year of just getting it," said Tulyani, who noted that Now Assist has removed manual tasks and helped Eaton engage with customers in forums and portals.
Eaton's AI agent goals include:
- Boosting productivity and becoming more efficient.
- Resolving tickets at a higher level with better customer service, resolution and speed.
- AI agents won't replace people, but for Eaton it's doing more with the same number of people.
- "We plan to expand and we can double our capacity over the next four years, but with our same headcount to support and drive the business," said Tulyani.
Eaton has started an AI council to leverage AI across the company. Tulyani noted that Eaton will use multiple models and plans to focus on use cases with ServiceNow being the core data structure, chassis and technical debt buster. "Our AI council is going to be evaluating which AIs are ready for us to integrate," he said.
Moving parts Tulyani said Eaton must navigate as it looks at AI agents include:
- "We see AI adopted well on the user side, but it was a little slower on productivity," said Tulyani. "We're focusing more on the productivity side with AI agents, but have issues and concerns."
- Those concerns include better data quality and Eaton has a project to clean up data and prep it for AI use cases.
- Eaton is looking for more automation to move data into the story and create structured data. The process is too manual today.
- AI will be used earlier in the data cleansing process to leverage AI later.
Tulyani didn't comment on Eaton's plans for managing consumption pricing as AI agents are adopted down the line. He did note that ServiceNow "was a little expensive," but worthwhile since it was delivering value and shedding tech debt.
