Many digital health companies focus on tracking movement and exercise, the “calories out” half of the weight loss equation. But efforts to track the other side, “calories in,” in any quantitative way, have proved elusive.

But while others try everything from spectrometry to blood sugar monitoring to help people automatically monitor their eating habits, Klue, a new startup that launched today at Health 2.0 in Santa Clara, California, is taking a simpler approach: Using gesture tracking to determine when a person is eating or drinking.

“There’s an enormous amount of valuable information encoded in our hand gestures about our behaviors,” CEO and cofounder Katelijn Vleugels told MobiHealthNews. “And today all those insights are basically untapped. So we developed an analytics technology that allows us to automatically, without any user intervention, detect and track behaviors, starting with eating and drinking.”

Vleugels was inspired by a mindfulness-focused health coach who helped her improve her eating habits by using a simple pebble placed next to her plate to remind her to think about what and how she was eating.

Klue isn’t launching a consumer product at the event, but it is coming out of stealth mode with an API, currently in private beta, that will allow other companies to use its gesture tracking technology with off-the-shelf wearables (for the demo at Health 2.0, an Apple Watch was used).

On the user interface side, Klue creates a “Consumption Graph” that lets people see when and how fast they’re eating. That data could help identify eating behaviors that sometimes lead to weight gain like eating too quickly or snacking in the middle of the day. And, via the API, that data could theoretically be tracked against other data like a manual food log or blood glucose data from a CGM.

“That consumption graph, we’re making that available to partners through an API. We’re hoping that they will integrate that with their products, their services, and their solutions,” Vleugels said. “Obviously companies focused on weight management [could benefit], but there’s also other applications: elderly care monitoring, being able to monitor if a loved one is drinking enough water through the day or staying well nourished, or for diabetes being able to correlate eating events with sugar levels. Those are just a couple of examples.”

The technology also allows for real-time responses as well as retroactive trend analysis. If a person is tracking calories in an app, for example, their Apple Watch could buzz every time they start eating to remind them to record it. These sorts of context-aware nudges are big in behavior change, and many believe they’re necessary for health apps to be truly effective.

“Having the combination of the data-driven insights and the delivery method which is real-time is something I’m really excited about,” Vleugels said.