PUTTING TOGETHER A SOLUTION
After the conference, we sat down as a team to synthesize everything — the seven interviews, the workshops, our conversations with FACO — into something we could actually build.
One thing came through clearly, confirming what we'd already started hearing in the interviews: reporting on ground activity is both important and genuinely hard to get right. Two things followed from that. On the ambassador side, whatever we built had to be simple — or at minimum, offer a simple option — because most ambassadors weren't going to adopt something complicated on top of everything else they already do. Voice kept coming up as the most natural way to deliver that.
The idea itself is straightforward. An ambassador finishes visiting a home, and instead of writing anything down or filling out a form, they just talk — the way they'd tell a friend what happened. What they saw, what they suggested, how the conversation went. That gets turned into the record a coordinator actually needs — the one with dates, categories, and the specific details a grant report asks for. No form. No app to learn. Just talking about something that already happened.
Simple to describe, harder to build well — especially given where this actually needs to work. Many of these communities barely have cell signal. Many ambassadors are older, retired, and not particularly interested in learning new software. Those constraints shaped everything: it had to work offline, be nearly impossible to get wrong, and fit how people actually talk.
On the coordinator side, the synthesis pointed somewhere different. FACO already has a rich, well-thought-out framework for how mitigation planning and ambassador programs fit together — the CWPP as the starting point, the menu of ambassador opportunities we'd learned about at the conference. Our job wasn't to invent something new. It was to reduce the friction of applying what FACO had already built — helping a coordinator start from their community's plan and land on a working ambassador menu, without reinventing that thinking each time. One more thread came out of the same synthesis, quieter than the others: how much ambassadors wanted to communicate with each other, share stories, and learn from other programs. That's still an open question for us, and may become its own feature down the line.
From there, we went back to FACO. Over a few working sessions, we walked through everything we'd synthesized and prioritized what to build first — shaping the roadmap we're working from now.
The voice idea became more than a feature along the way. In a conversation with an AI advisor about what it would actually take to build — something that listens to an ambassador talk and reliably turns that into a usable record, offline, on an ordinary phone — it became clear nobody had solved this yet. It was a real research question, not just an engineering one. That's what led us to apply for another NSF grant, this time built specifically around it.
We've since started prototyping — rough screens for a coordinator setting up a new program, and for an ambassador logging what they did by talking. Nothing final, but enough to put in front of the people who'd first talked to us, and ask honestly: does this look like something you'd use?
That's where things stand now. We're in a second round of research, showing the prototype to coordinators — and what we're hearing back is already reshaping our thinking.
Next: what happened when we brought the prototype back to the people who inspired it.