August, 2026
With 10+ years in design, I've been focusing on shipping production-ready AI experiences, end-to-end from discovery to code.
VERIZON CONNECT – 2025
I led design end to end for an AI insights feature that shipped to 30,000 fleet managers. I defined what an insight says, how it earns attention, and how users tell us when the system gets things wrong. By the end of a six-month beta, 93% of the group called it essential to their work.
Getting an answer out of Reveal was manual. Users ran several reports, exported CSVs and rebuilt them in a spreadsheet, often ending up with something that did not answer the original question. Most gave up before that point.
Verizon Connect also needed a credible AI feature to hold its ground against newer fleet products.
An AI agent reads vehicle data over 7 and 30 day windows, compares each vehicle against the fleet average and its own history, and writes a plain-language insight when it finds something worth acting on.
The decision that carried the feature was the reading order. Every card leads with the operational area, the size and direction of the change, then one sentence explaining why it matters. An user scans the feed in under a minute and opens only what needs a decision.
Beta users called it essential. Across six months of testing, 93% of the 300 people using it described the tool as essential to their work.
One in seven came back in the first week. Of the 30,000 fleet managers it launched to, 14% opened the feature and returned at least twice.