How can we help technicians to work with — and trust — an AI system?

Product design • Myrspoven
Project Overview
Myrspoven is a pioneer in AI-driven building optimization. Their AI, which helps save buildings save energy by optimizing HVAC systems, was already running in a significant number of buildings, but the tools the team had created to let customers manage that AI were still pretty bare-bones, leading to a high support load. The team had a vision of a customer portal that would let users crunch big data to find savings and help them understand their buildings better, while also giving a peek into the black box of the AI. 

The project is ongoing and currently in private beta. A full case study is coming soon.
My Contributions
UX research
Wireframes and prototypes
Product design
Visual design
Design system
Managing frontend development
Usability testing
Visiting a client's boiler room and HVAC facilities to help research context of use
An interface uses the"small multiples" data visualization pattern to let technician users quickly scan graphs of their HVAC system components for problem areas.
A selection of built and proposed screens from the app
Observations from the first round of user tests were affinity mapped, then translated into to-dos for app improvements
A proposed revision of the building dashboard after user testing that focuses more on actionable metrics