There's a version of AI in software that everyone knows and nobody asked for. A feature appears in your platform one morning, uninvited. A sparkle icon shows up next to a button you have used for years. Suddenly you're wondering what it does with your data and why nobody asked whether you wanted it.
That's not how we do it at Via.
This spring, we ran a series of Via Labs sessions, live working conversations where international educators sat down with our product team and helped shape what AI in Via should actually look like. Not a demo. Not a sales pitch. A prototype on the screen, a product manager taking notes, and practitioners saying "yes, that," "no, not that," and "have you thought about this?"
Here's what that process looks like from the inside, and why we think it's the only honest way to build AI for international education.
Our first Labs session in April was simple: virtual sticky notes and one question. Where does your day lose time to finding, piecing together, and re-checking information?
Two ideas rose to the top of that conversation in an overwhelming way. So our product and UX team did what you'd hope they'd do: they built rough, working prototypes of both and brought them back to the same community for round two. Nothing polished. Nothing final. As our product manager put it in the session, the point is to "give you a kernel of an idea" and let real practitioners bend it into shape.
That second conversation is where the ideas got sharper. Study abroad advisors told us exactly which flags matter before a student walks into their office. Travel risk and safety teams told us where trusted data has to come from. And when something didn't land, they told us that too. That feedback changes what gets built. That's the entire point.
1. Opt-in, always. Every AI feature Via puts out will be something you choose. If you want it, it's there. If you don't, your Via works exactly the way it does today. No forced rollouts, no surprise sparkle icons.
2. Your data stays in Via. The concepts we're exploring don't ship your student information off to external AI services. They work with the data your institution already puts into Via, surfaced in one place instead of five tabs. When a feature touches something sensitive, it pulls from trusted sources you already rely on. You can read how we handle data security and privacy in Via.
3. Internal-facing first. Anything we build in this space starts as a tool for you and your team, not your students. Educators in our sessions were clear: with this much nuance across program types and student populations, humans stay in the loop. We agree. Good stewardship comes before wide rollout.
Because the alternative is spending months building something that doesn't land. Because the people advising students, managing travel risk, and pulling the weekly reports know things a product roadmap never will. And because trust, once lost, is very hard to win back in a community as connected as international education.
The features we prototyped are not on a release schedule. There's no pricing page hiding behind them. They're ideas being shaped in the open with the people who would actually use them, and some of what we heard is already changing how we think.
Curious what international educators actually asked for when we handed them the whiteboard? That's the next post in this series, and some of their requests might sound very familiar. 👀