Bearpaw

One of the things I try to incorporate into my personal AI design work is hardware integration. One of the first was based on a hobby need.

I kept an old Windows laptop around to run amateur radio apps that were Windows only. One day, it came up with a boot error, so I needed Mac versions of some of the apps I used. In particular, I needed a way to program my Bearcat BC125AT scanner. Uniden’s stock software was Windows-only, so I took this dilemma as a way to handle product design from start to finish while experimenting with AI coding and interfacing with the scanner hardware.

The MVP

At first, my goal was only to make a replacement for that programming software. It only took a session or two to complete that, so I asked myself, “What else would I want in scanner software that wasn’t in the stock app?”

I knew that it was possible in a separate app on Windows to control the scanner, so I wanted to incorporate that. I wanted logging of radio receptions and a report of what frequencies are busiest. I also wanted compatibility with the stock programming files. Finally, I wanted a well-designed app that had the same attention to user experience that I gave to my time building Redwood at Oracle.

After about eight months of off and on tinkering, it was to a point that felt complete to me. I realized there were undoubtedly other radio hobbyists that would benefit from this. Based on the way it was being built, it was easy to offer a cross-platform app. I set up a website and posted the link on Reddit and on a highly active radio forum.

I quickly got positive feedback from users. I also quickly got requests for Bearpaw to support other Uniden scanners. The next question was, “How many Uniden scanners can I support?” I had a second scanner, so I added support for that. With two scanners supported, the requests came for several other popular models.

Remembering the user

I realized that the temptation to make Bearpaw a kitchen-sink app was possible. I decided I would stick with doing few things but doing them exceptionally well. I also needed to remember that I have real users now. So, I made sure to keep up with online feedback and add only new features or settings that offer an exceptional user experience. Every suggestion so far has been a positive addition to the usability of Bearpaw. And, I have a roadmap of features to come. All user-suggested and in line with my design principle.

Future versions

Since those first posts in late August, Bearpaw has been downloaded over 1,000 times. And, funnily enough, half of its user base is on Windows. Some users have valued it enough to pay me.

I’ve made the goal of supporting every kind of Uniden scanner that there is. The problem is I need to own those scanners to develop against and use for QA testing. That’s going to be incredibly expensive, since these scanners cost $100-$950 each. Any payment that I receive for this app is going back into the purchase of scanners to be supported. I’m also working on ways to borrow the more expensive scanners for development.

I just got a third scanner off eBay, and it’s the next scanner Bearpaw will fully support.