T-Mobile · Agentic AI

NetHive IQ

Designing AI for complex network operations — without taking control away from the people doing the work.

next project

Overview

How do you introduce AI into complex operational workflows without taking control away from the people doing the work?

I led UX strategy and agentic AI exploration for a 6-week rapid product evaluation, working with an 8-member team to explore how AI could support network professionals across different roles and workflows.

The work resulted in a persona-based framework for human + AI collaboration, mapping user needs, workflows, agent behavior, and safeguards.

Role

Senior UX Design Lead

Team

8-member cross-functional team

Timeline

6-week rapid product evaluation

Focus

  • Strategy & Research
  • User Workflows
  • Agentic AI Design Exploration

Chapter 01

Understand the Users

NetHive IQ served users with very different responsibilities, expertise, and goals.

We used personas to understand where network executives, operations teams, and field engineers needed different levels of information, guidance, and AI assistance.

Network Executives

Decisions

Operations Teams

Time

Field Engineers

Tasks

Different users. Different needs. One AI experience.

The challenge was not simply to introduce AI. It was to understand how AI should behave differently depending on the person, their expertise, and what they were trying to accomplish.

  1. Persona
  2. Expertise
  3. Context
  4. Intent
  5. Response

Chapter 02

Map Human + AI Workflows

We mapped key user workflows to understand where AI could reduce friction, provide context, and help people move from insight to action.

The goal wasn't to replace existing workflows with AI.

It was to understand where AI could become a useful partner within them.

Human + AI collaboration

  1. AI suggests
  2. User reviews
  3. User decides
  4. AI assists
  5. User acts

The human remains in control at every step.

Designing for trust

We explored safeguards around critical AI interactions so users could understand what the AI was recommending, why it was making a recommendation, and when human confirmation was required.

  1. Assist
  2. Explain
  3. Recommend
  4. Confirm
  5. Act

AI-powered onboarding and learning

The AI experience extended beyond chat.

We explored how an agent could help new and returning users understand the platform, discover relevant modules, and continue learning within the context of their work.

New User

  1. First Login
  2. Guided Onboarding
  3. AI Conversation
  4. Understand Goal
  5. Recommend Modules
  6. Request Access
  7. Confirmation

Returning User

  1. Dashboard
  2. Relevant Modules
  3. Continue Learning
  4. Ask AI for Help
  5. Contextual Guidance
  6. Continue Workflow

Chapter 03

Design the Agentic Experience

With the users and workflows mapped, we explored practical ways to bring AI into the experience.

The goal was to move AI from a standalone feature toward a workflow enabler.

Insight → Decision → Action

AI should help people move through complex work — not create another destination they have to manage.

AI Experience Concepts

What's Next?

Context-aware suggestions that help users move from conversation toward relevant actions.

Contextual Onboarding

An adaptive AI conversation that learns what the user needs and recommends relevant experiences.

Saved & Recent Prompts

Support repeat workflows by allowing users to save and quickly return to useful prompts.

Persistent AI Assistance

Allow users to move between conversation and product screens while maintaining context.