Predictive Support Hub

A support platform that reduced Average Handling Time by 75% by consolidating information scattered across 32 systems into a contextual workspace.

  • 12 min → 3 min AHT
  • 4× potential capacity
Role
Product Designer
Timeline
9 weeks
Scope
Support operations, field research, AI-assisted workflows, legacy system consolidation
Outcome
12 min → 3 min average handling time, 32 tools → 1 workspace

Overview

A large telecommunications provider wanted to reduce Average Handling Time (AHT) across its support operation.

Support agents relied on 32 different systems to diagnose issues, access customer information and resolve requests. Over time, experienced agents developed their own shortcuts, notes and workflows to navigate this complexity. Newer agents often depended on them to understand where information lived and which tools to use.

Over a 9-week engagement, I led research across support operations in New York and Texas, alongside remote interviews with agents in Egypt and Jamaica, helping define and design a contextual workspace that reduced Average Handling Time from 12 minutes to 3 minutes.

Understanding the Operation

Over two weeks, we conducted contextual interviews and shadowing sessions with support agents across different levels of experience in New York and Texas, alongside remote sessions with teams in Egypt and Jamaica.

We observed live support calls, mapped workflows and documented how agents moved between systems during customer interactions.

Although 32 systems were available, agents relied on only 7 of them regularly. Most support requests fell into a small number of recurring categories, including billing issues, network problems and plan changes.

The information needed to solve those issues already existed across the organization. The effort came from locating it, validating it and assembling enough context to help the customer.

Findings were later reviewed with experienced agents through collaborative workshops, helping prioritize information and workflows for the first release.

Designing the Workspace

The solution centered on a contextual 360° customer view.

As a call arrived, the platform assembled relevant information automatically, including customer details, billing status, network indicators, support history and recommended next actions.

Instead of spending the first minutes of a conversation gathering context across multiple systems, agents could begin with a complete picture of the situation.

360° customer workspace showing services, devices, network panorama, billing, order history, and active call tools in one view

The platform also incorporated an internal recommendation engine that surfaced troubleshooting guidance and suggested actions based on the customer’s context.

Suggested Interaction prompting the agent about a past-due balance with pay now and decline actions

For technical issues, the same pattern applied: the workspace surfaced the next step, then opened the specialized tool without leaving the call context.

360° view with a suggested interaction directing the agent to check connected devices in Panorama after a modem reset
Panorama device view inside the same workspace, with network topology, quick actions, and a reset modem suggestion

Working Within Constraints

Replacing all 32 systems would have required a large migration effort and significantly increased implementation time.

We focused instead on integrating the systems agents used most frequently while providing contextual access to the remaining tools when needed.

This approach reduced implementation risk, accelerated delivery and allowed the organization to validate the concept in production without waiting for a broader transformation program.

Results

12 min → 3 min AHT

75% reduction in Average Handling Time

4× increase in potential handling capacity

Additional outcomes:

  • Reduced dependency on specialist agents
  • Faster onboarding for new team members
  • Less time spent navigating between systems
  • More time focused on customer issues

The platform was released through a pilot program with experienced agents before broader adoption.

Reflection

The project revealed how much operational knowledge had become tied to individual experience.

Experienced agents had learned where information lived, which systems mattered and how to move between them efficiently. Bringing that context together in a single workspace reduced the effort required to perform the job and made the operation less dependent on institutional knowledge.

The most valuable outcome was not making a new interface, but making the existing information easier to use.