ServiceNow is building a data foundation to support the use of artificial intelligence (AI) agents, solving what one of its executives deemed as “data hell”, the single biggest roadblock to successful enterprise AI adoption.
“There’s a very sobering statistic from Gartner that says that 60% of AI projects are likely to fail by 2026 because their data is not AI-ready,” said Gaurav Rewari, senior vice-president and general manager of data and analytics products at ServiceNow. “That prompted us to think very critically about how we could support our customers through their journey to deploy agentic AI at scale.”
To address this, ServiceNow has built three core components. The first is RaptorDB, a new AI-native database designed to handle the massive increase in interactions from AI agents. “We need an underlying AI database that can execute these workflows super-fast,” Rewari said, noting it must support both operational and analytical workloads as AI agents think, reason and then act.
The second piece is Workflow Data Fabric, which allows ServiceNow’s AI agents to connect to and understand data from external sources like Snowflake, Databricks and Oracle. “We need to give these AI agents an education in data outside our four walls,” Rewari explained.
The final component is the company’s recent acquisition of Data.world, a data catalogue that helps agents understand the meaning, context and lineage of data across the enterprise. “Think of it as a next-generation data dictionary plus encyclopaedia all rolled into one,” Rewari said. “It’s everything you ever wanted to know about a data field but were too afraid to ask.”
As other enterprise software giants such as Salesforce make their own data-centric plays, including its recent intent to acquire Informatica, Rewari noted that ServiceNow’s advantage lies in its unified architecture.
“ServiceNow has taken a very disciplined approach to organic development, and when we make inorganic moves like acquisitions, we re-platform it back, maintaining the purity and unity of a single platform and data model,” he said. This, Rewari claims, is a key differentiator from competitors whose capabilities may be spread across multiple acquisitions and platforms.
ServiceNow has taken a very disciplined approach to organic development, and when we make inorganic moves like acquisitions, we re-platform it back, maintaining the purity and unity of a single platform and data model
Gaurav Rewari, ServiceNow
For CIOs managing a heterogeneous environment, ServiceNow aims to provide a centralised management layer. The company’s AI Control Tower is designed to offer visibility and governance over both its own agents and those from third-party providers like Google or Microsoft.
This control tower is central to managing the spectrum of agent autonomy, a key concern in areas like governance, risk and compliance (GRC) and security. It allows leaders to “continuously measure how you’re doing in terms of level of automation,” Rewari said. “You can keep dialling that up until such time that you’re completely comfortable with letting it out in the wild.”
BI ambitions
While the immediate focus is on powering agents for workflow automation, ServiceNow has broader ambitions in the business intelligence (BI) and analytics space.
“Because of our roots as a workflow orchestration company, we feel we’re uniquely positioned to deliver what we call insight-to-action workflows,” Rewari said. He contrasted this with data-native companies for whom the “take action piece is new to them.”
Looking ahead, Rewari said CIOs should expect ServiceNow to invest heavily in its analytics offerings, expanding into predictive and prescriptive analytics and scenario modelling. This signals the company’s intention to challenge traditional BI players, with a focus on natural language and conversational interfaces.
“We think there’s an opportunity for AI to transform decision support,” Rewari said. “The paradigm with which you will get insights will be conversational, with Gartner predicting that 60% of companies will retire dashboards and find new ways of getting their questions answered. We think we could be very relevant in that world, so stay tuned.”
Read more about IT in APAC
Many generative AI projects are failing to scale because firms are overlooking foundational needs like network modernisation and governance frameworks, says a regional leader at NTT Data.
As organisations move from AI hype to reality, a decline in trust for AI outputs is not a sign of failure, but a signal of market maturity, according to Bhavya Kapoor, Avanade’s APAC president.
Datadog is executing a multi-year growth plan for Asia-Pacific and Japan, focusing on data residency, localisation and AI-driven observability to grow its market share.
Read more on IT operations management and IT support