Node-Based Coaching Automation Tool

Design Concept

Node-Based Coaching Automation Tool

Problem Statement

Call center managers often struggle to coach agents effectively on performance metrics. Manual interventions are inconsistent, reactive, and time-consuming. NiCE needed a modular automation system that would allow managers to build dynamic coaching flows tailored to agent performance, without requiring technical expertise.

Goals

• Design a visual, modular system for building automated coaching flow
• Make it intuitive for non-technical users to assign agents, define objectives, and select outcomes
• Support complex logic while maintaining clarity and usability
• Lay the groundwork for future AI integration to assist in building workflows based on historical data and best practices
• Provide realistic templates that mimic common coaching scenarios to reduce setup time and improve adoption

Discovery & Decision Making

Rather than conducting user interviews, I approached this project through a systems-first lens. I researched existing node-based editors to understand how complex logic could be made visually intuitive. One of the most influential references was Blender’s node-based material editor, which elegantly balances modularity, clarity, and flexibility.

I analyzed how nodes are structured, how connections are formed, and how users navigate branching logic. This helped inform the architecture of our coaching flow builder, where each module represents a distinct action, condition, or objective.

Importantly, the system was designed to support multiple performance metrics (e.g., FCR, CSAT, QA compliance) and a wide range of outcomes. This flexibility was central to the design, ensuring managers could tailor flows to their team’s needs without being locked into a single metric or coaching style.

Solution

• Design a drag-and-drop canvas for building flows with zoom and pan features
• Create modular components with editable fields and dropdowns
• Enable branching logic based on performance ranges
• Develope starter templates for common scenarios like FCR improvement
• Structure the system to support future AI features, such as auto-suggested flows and smart branching logic
• Add loop-back functionality to reassign agents if performance stagnates

Interface Architecture

• Start Module: Names the flow and assigns agents or teams
• Objective Module: Selects a performance metric (e.g., FCR) and defines ranges:
• Bad: FCR < 60%
• Average: FCR 60–80%
• Good: FCR > 80%
• Each range links to a Scenario Set, which connects to an Outcome Module:
• Bad: Triggers a coaching session and assigns a workaid
• Average: Sends a notification to the supervisor and assigns a gamification challenge
• Good: Awards a badge and sends a congratulatory message

Outcome modules can connect to another objective or loop back to the start if performance doesn’t improve, creating a continuous feedback loop.

Anticipated Impact

While the product had not entered development, the design was created with several key outcomes in mind:

• Streamline coaching workflows by allowing managers to automate performance interventions across multiple metrics
• Reduce manual effort and inconsistency in coaching by providing structured, repeatable flows
• Improve agent performance through timely, targeted feedback and gamified incentives
• Increase adoption by offering realistic templates that mirror common coaching scenarios
• Lay the foundation for AI-driven enhancements, enabling future features like auto-suggested flows and smart branching logic
• Enhance visibility and accountability for supervisors managing large teams

What I Learned

This project deepened my understanding of how modular design can simplify complex workflows. Designing for non-technical users required empathy, clarity, and iterative testing. I learned how to balance flexibility with structure, giving users control while guiding them toward best practices. Planning for AI integration challenged me to think ahead and design with extensibility in mind. Most importantly, I saw how thoughtful UX can empower managers to coach more effectively and improve agent performance at scale.