The most recent use case has been with an internally developed boostrapped device that will focus on employee tracking and workflow management. We expect to launch this in Q1 of 2019 for our first real world implementations combined with a SaaS offering.
Another was for a B2C device that was developed for a private company to provide service and interactivity to their customers with a mobile App (controlling and monitoring the device) in a very niche space.
Happy to talk about specifics - feel free to reach out via email (in profile).
Project delayed because AT gpio is pretty slow, and couldn't pilot the matrix fast enough. We're then using a STM32, and we went down the rabbit hole of creating our own rpi hat. Lot of learnings.
If you're not building an IoT device, why? Don't know electronics? Too complex? Too costly?