Object Model Re-architecture

Overview

PagerDuty's core object model—how services, escalation policies, and schedules relate to each other—had grown organically and no longer matched how customers thought about their operations. I led the effort to redesign this fundamental architecture.

The Problem

User research revealed that customers had a mental model of PagerDuty that didn't match how the product actually worked. This disconnect caused confusion, led to misconfiguration, and limited what features we could build. The existing model was also blocking machine learning initiatives that could help customers respond to incidents faster.

The Solution

I led extensive user research to understand how customers actually thought about their incident response workflows. Working closely with engineers, we designed a new object model that matched customer mental models while maintaining backward compatibility. The migration was designed to be seamless—existing customers saw no disruption to their workflows.

Results

• Released with no disruption to existing customer workflows
• Enabled new machine learning functionality to be built on the platform
• Contributed to the PagerDuty IPO in 2019
• Reduced customer confusion and support tickets related to configuration