Downloads: 0
Research Paper | Computer Science and Information Technology | Volume 15 Issue 5, May 2026 | Pages: 1001 - 1008 | United States
Adaptive Role-Based UI Personalization in ServiceNow Configurable Workspace: A Reinforcement Learning Approach to Agent Productivity Optimization
Abstract: ServiceNow Configurable Workspace presents agents with a rich but static UI environment in which panel layouts, widget configurations, and information hierarchies are fixed at design time by platform administrators. This one-size-fits-all approach fails to account for the substantial variation in task patterns, information consumption behaviors, and interaction preferences across agents within the same role - variation that directly affects task completion efficiency and cognitive load. This paper proposes RLWS (Reinforcement Learning Workspace Shaper), a novel framework that applies contextual multi-armed bandit reinforcement learning to dynamically adapt ServiceNow Configurable Workspace layouts at the individual agent level, optimizing UI configurations based on continuous feedback signals derived from task completion time, click-through patterns, scroll depth, widget interaction frequency, and SLA compliance rates. RLWS operates within ServiceNow's UI Builder framework through a custom workspace component that intercepts agent interaction events and applies layout recommendations via the workspace's dynamic configuration API. Evaluation across 214 agents in three enterprise ServiceNow deployments over sixteen weeks demonstrates that RLWS-optimized workspaces reduce mean task completion time by 22%, decrease scroll distance per task by 41%, improve SLA compliance rates by 8.4 percentage points, and achieve agent satisfaction scores of 4.3/5.0 compared to 3.6/5.0 for static layouts. A multi-armed bandit algorithm selection framework, a reward function design guide, and a ServiceNow UI Builder integration architecture are presented.
Keywords: Reinforcement Learning, ServiceNow Configurable Workspace, UI Personalization, Multi-Armed Bandit, Thompson Sampling, Agent Productivity, Cognitive Load
How to Cite?: Abhinav Reddy Pullikallu, "Adaptive Role-Based UI Personalization in ServiceNow Configurable Workspace: A Reinforcement Learning Approach to Agent Productivity Optimization", Volume 15 Issue 5, May 2026, International Journal of Science and Research (IJSR), Pages: 1001-1008, https://www.ijsr.net/getabstract.php?paperid=SR26515093450, DOI: https://dx.dx.doi.org/10.21275/SR26515093450