AI + workforce management

AI Workforce Assistant

Designing AI assistance that helps managers act, not just find information.

An AI-assisted workforce experience that helps frontline managers understand complex situations, evaluate recommendations, and act, with human judgment kept in control.

Challenge

Reduce the cognitive effort of everyday workforce decisions without removing manager judgment.

My role

Principal Product Designer

Product strategy · UX research · Interaction design · AI experience design · Prototyping · Validation

Company
UKG
Focus
Human-AI interaction · Responsible AI
Scope
Enterprise workforce management

The challenge

Workforce decisions required too much cognitive effort.

Frontline managers moved between scheduling, attendance, staffing, and employee-management tools to understand what was happening and decide what to do next.

The opportunity wasn't to add AI. It was to reduce the complexity of these decisions without removing manager judgment or surfacing recommendations managers couldn't trust.

My role

Defined the experience model for AI-assisted workforce management.

Partnered with Product and Engineering to define the product vision, conversational interaction model, AI-assisted workflows, and responsible-AI principles. Led research, interaction design, prototyping, and validation.

  • Product strategy
  • UX research
  • Interaction design
  • AI experience design
  • Prototyping
  • Validation

Key decisions

Designing AI around the decision, not the technology.

01

Start with the decision, not the AI capability.

Instead of asking where AI could be added, we started with the workforce decisions managers struggled to make and the context they needed to make them confidently. Assistance was designed around those moments of judgment rather than treated as a feature looking for a use case.

Assistance embedded in the everyday workspace, not a separate destination.

02

Integrate AI into existing workflows.

Rather than creating a separate AI destination, we explored how assistance could appear within the workflows managers already used. Guidance could surface at the moment a decision needed to be made, reducing the need to leave the task or search across multiple experiences for context.

An interactive prototype used to test the conversation flow with users.

03

Make recommendations explainable and actionable.

Guidance needed to do more than recommend an action. A “How Bryte got this answer” view showed what required attention, why it mattered, and the supporting context behind the answer — the intent inferred, the agents consulted, and the sources used. Managers could judge the fit before deciding, and the final decision stayed with them.

Full product context: the assistant sits beside the manager's workspace.
Detail: the reasoning behind the answer, with the decision left to the manager.

The solution

An AI-assisted workspace for everyday workforce decisions.

The resulting experience brought workforce context, guidance, explanations, and actions into a shared decision workspace. Managers could understand what required attention, explore why it mattered, review supporting information, and decide what action to take — while remaining in control of the final decision.

01

Decision workspace

Relevant workforce information, guidance, and actions come together around the decision the manager needs to make.

An answer in context, with a handoff to the destination that completes it.
  • Assistance appears within the workflow, not in a separate destination.
  • The manager's current workforce situation frames the conversation.
Detail: guidance connects to the next step, taken by the manager.

02

Explainable guidance

Supporting context helps managers understand why something needs attention before deciding what action to take.

Detail: the assistant explains how it reached the answer.

Impact

Creating a foundation for responsible AI-assisted workforce experiences.

Validated

AI assistance directly within manager workflows.

Simplified

Complex workforce information into guided decisions.

Aligned

Product, design, and engineering around an AI interaction model.

Foundation

Reusable patterns for explainable, human-controlled AI.

The goal wasn't to automate the manager. It was to give managers better context, clearer guidance, and faster ways to act, while keeping the decision in their hands.