NextEra Energy

The challenge

What was keeping the platform from being ready?

My research focused on three questions:

1. Where did sales representatives struggle with the existing platform?
2. What support did they need throughout the sales process?
3. What would make the platform easier to use during customer conversations and independently?

Research approach

Understanding the work before evaluating the screens

To better understand sales representatives’ needs, I combined interviews, usability testing, and an interface review to identify where the platform fell short.

Recruit sales representatives

Understand their workflow

Observe platform use

Evaluate the interface

Share findings with stakeholders

Seeing the sales journey beyond the platform

I mapped the sales representative’s journey from researching potential clients to advising customers and maintaining relationships. Bringing the interview findings into one view helped reveal where scattered information, internal coordination, and ongoing customer questions made the process difficult.

The map connected these challenges to specific stages of the sales journey, showing where the platform could better support representatives’ work.

Key findings

What did sales representatives actually need?

Sales representatives relied on multiple tools and sources to build an understanding of each client. Bringing that information together and making sense of it added work before they could advise a customer.

“We're having to gather info from a lot of different tools — and [platform] could be better, an easier way to find those basic things.” - NextEra Energy Sales Representative

The opportunity extended beyond displaying energy data. The platform could help representatives find relevant client information and understand it in one place.

Design direction: A dedicated client page bringing together relevant information and sources, with access that preserves confidentiality.

Our initial focus was the website, but interviews revealed friction across the broader sales process. Representatives depended on input from teams such as legal, credit, and pricing. Conflicting priorities and internal coordination could delay progress for clients.

“Getting the information isn't the difficult part. It's analyzing it and being aligned internally [that] seems to be the difficult part sometimes.” - NextEra Energy Sales Representative

This finding expanded how I understood the problem. Some delays came from coordination between people and teams, so interface changes alone would not address every pain point.

Design direction: Explore how shared client information could support coordination, while identifying which issues would require process changes beyond the platform.

Representatives wanted customers to understand their data and find answers without always needing a one-on-one conversation. Yet representatives themselves encountered hidden features and difficult navigation, and some were less comfortable using digital tools.

These findings raised concerns about independent customer use. They also identified questions that would need to be tested directly with customers.

Design direction: Create a clearer customer homepage, standardize filters, improve graph contrast, and add help where questions are likely to arise.

The platform depended on customer-specific data, but representatives were unsure how to use it with new clients whose information was not yet available.

This exposed a gap in the experience: the platform needed to support the steps before a complete customer view existed.

Design direction: Define onboarding and clarify what information is required, what is missing, and what users can do next.

Recommendations

How could these findings guide the redesign?

Make the interface easier to understand

Standardize filters, improve graph contrast, add contextual help, and use validation indicators to clarify feedback.

Create clearer entry points

Develop a customer homepage and define onboarding for clients whose data is not yet available.

Bring client information together

Explore a shared client page that supports access to relevant sources while preserving confidentiality.

Explore support for customer questions

Consider AI-assisted help as a potential way to answer questions without requiring a sales representative for every interaction.

Together, these findings pointed toward a platform that makes information easier to find, understand, and act on — for both sales representatives and customers.

The recommendations focused on clearer navigation, smoother onboarding, shared client information, and more independent customer support, with further testing needed to refine these ideas.

Handoff and next steps

What did the research contribute?

I shared my findings and recommendations with stakeholders to help guide the redesign, connecting usability issues with the challenges sales representatives faced in their daily work.

While my internship ended before design work began, the proposed next step was to agree on priorities and build a roadmap before testing changes with sales representatives and customers.

What did I learn?

How to connect user needs with business realities

This project was SO cool and really taught me to look beyond the initial brief. What started as a website review turned into a deeper look at how sales teams work, from piecing together information across tools to coordinating with different departments. Interviews uncovered so much I wouldn’t have noticed by looking at the screens alone.

I also loved the challenge of making research click for people who don’t work in UX. It pushed me to turn a lot of information into clear takeaways and recommendations the team could use.

If I could keep going, I’d be excited to test those ideas with sales representatives and customers, especially to see how they make sense of energy data on their own and get started when some information is still missing.

The Problem

Heuristic evaluations are an extremely time-consuming process

After spending six hours on a heuristic evaluation, I asked myself: “how much time do other designers spend on the same process?”

Casual conversations with my HumanITy teammates revealed a shared frustration: these reviews were valuable, but time-consuming. That sparked my interest in using AI to speed up the process while keeping designers’ judgment at the center.

Research

I interviewed eight designers to understand how they approached heuristic evaluations and where the process became tedious. I focused on what they would expect from an AI-generated baseline and, most importantly, how an agent could support their judgment without replacing it.

The interviews revealed four priorities for building an AI agent that would be useful to my team:

Implementation

Using the interview findings, I compared the AI platforms available at NextEra and chose Gemini for its reasoning and ability to analyze detailed content. I then built the agent and tested it on the same four screens I had reviewed manually.

It generated baseline heuristic scores in just seconds — a starting point for designers to review, compared with the six hours I spent on the full manual evaluation.

I gathered feedback through designer surveys and used it to refine the agent. The team continues to use it and reports saving hours on heuristic evaluations!

What did I learn?

How to keep users at the center, even when building with AI

It was really rewarding to build something my team could use, and I had so much fun exploring AI along the way. I applied the research skills I developed on the C&I project while independently comparing platforms and shaping the agent around designers’ needs.

This experience made me really excited to keep finding ways AI can make everyday work easier while keeping people at the center.