Reimagining Workflows in the Agentic Age
The use of agentic AI has grown dramatically in the past few years. According to Kumar, Box has seen the use of Model Context Protocol (MCP) increase by a factor of 30 among its customers, “showing that they need to access information from numerous silos, but they also need to do it securely.” Another use that Samsung makes of Box is to unify data for over 260,000 global employees. Box centralized the data and categorized any data with sensitive or personally identifiable information so it could be governed properly.
Implementing agentic AI can be a challenge due to lack of expertise within the company. “There is a lot that can be done with agentic workflows, but many companies don’t have the ability to generate the eval sets to measure the reliability of the agent,” Kumar commented. “Box can build out eval sets to test the accuracy and effectiveness of their agentic workflows.” Finally, change management is always a factor. “We work closely with customers to structure pilot studies that provide a good ROI experience, then work with them to implement workflows more broadly,” he added.
Personalizing Customer Interactions
BPM systems found strong adoption in highly regulated industries such as financial services and healthcare that have complex processes. Automating these processes provides consistency and auditability, and the addition of AI workflow provides agility. NatWest Group plc, a British banking and insurance holding company based in Edinburgh, Scotland, has a corporate pedigree that dates back several centuries. With over 20 million customers in retail, commercial, and private banking, it had outgrown its on-prem cloud that was supporting the 40 engagement channels that supported personalized account services.
NatWest looked to Pega to upgrade its infrastructure by migrating from its private cloud to the Pega Cloud to improve system stability and performance. The digital transformation was planned in stages to minimize disruption, using an approach for each of its customer sectors that matched its needs. This approach ranged from fully cloud native to modernizing existing infrastructure to adopting a hybrid approach.
“Legacy systems and legacy culture are obstacles to implementing enterprise workflow,” said Matt Healy, senior director of intelligent automation at Pega. “Providing a foundation is essential.” The upgrades allowed NatWest to connect numerous digital and physical channels, offering its customers a much more unified experience. Pega supports 3.6 billion personalized interactions per year, and 60% of all sales are prompted by Pega next best action (NBA) recommendations.
Real-time data streaming allows for more rapid customer decision making, which provides an improved workflow for the customer lifecycle. “Traditional marketing automation is not effective anymore,” Healy contended. “Campaign-based marketing needs to move to one-on-one interactions.” Pega’s 1:1 Operations Manager orchestrates personalized customer actions across channels to coordinate the customer journey across many channels.
Decisions made in real time provide appropriate offers and messaging for customers. “If the customer does not respond positively, then the adaptive AI can immediately help the brand pivot to more personalized content,” said Tara DeZao, senior director of 1:1 customer engagement. The Customer Decision Hub (CDH) is the source of the decisions. It analyzes customer signals and recommends an NBA. “The decision is made in just 200 milliseconds,” noted DeZao, “allowing for a smooth customer experience.”