NextEra Energy · Sr. UX Designer / Product Lead · 2024–2025
Multibillion-dollar decisions deserve better than a spreadsheet.
NextEra's pricing team was running $18B in revenue decisions through Excel and institutional knowledge. The brief was to bring AI into the workflow without taking the analyst out of it. Not a black box, not a spreadsheet, something between. I led the research, designed the transparency layer, and built the patterns that became the platform's standard for human-AI decision UX.
$18B
annual revenue supported
12+
structured user sessions
AI
cohort prioritization patterns
The problem
Pricing analysts at a top-five US energy company were making multibillion-dollar contract calls in Excel. The institutional knowledge was real and earned, and it also lived in a couple of analysts' heads, which is a single point of failure measured in billions. The team wanted AI in the loop, but they had seen enough black-box products to know that an AI tool experts can't audit is an AI tool experts will route around.
Research and design
Twelve-plus structured sessions with analysts and traders. The recurring theme was trust calibration. Experts will use AI when they can see how it got there. They won't use it when they can't, and they really won't use it when it's confident in a way the data doesn't support. So the design problem wasn't "show the recommendation." It was "show the recommendation, the confidence, the uncertainty, the reasoning, and the override path," all without burying the answer.
Experts will use AI when they can see how it got there. The design job is making the how visible without burying the what.
The patterns
Built reusable patterns for human-AI collaborative decision-making. Confidence shown as range, not point. Reasoning available on demand without dominating the screen. Override paths first-class, not buried. Cohort prioritization that surfaces the rationale alongside the ranking. These weren't novel ideas. They were the boring, durable ones, and they shipped.
Outcome
Tool launched and was adopted by the core pricing team. The patterns were documented and handed off as reusable foundations for future AI-assisted product work across the platform. The deeper win is harder to measure. A pricing team that trusts its tools makes better calls, and the tools they trust are the ones designed for the way experts actually think.
Next case study
Banner Engineering
Banner Engineering — Unified UX Framework