Photos play a huge role in helping eaters decide what to order. This past year, our team focused on collecting and surfacing user-generated dish images for menu items without merchant photography. You may have seen Uber Cash offers encouraging you to share photos of your meals!

As we collected more images, we noticed that some didn’t quite capture how good the food really looks. Photos were sometimes a bit blurry, tightly cropped, or taken in low light. We saw an opportunity to incorporate intuitive guidance into the in-app camera to help users feel confident in their photos.

So we set out to design a photo-taking experience that feels effortless and instantly validating.

The result was a real-time guided camera that translates common rejection reasons (like framing and lighting) into gentle, clear feedback. To help users who typically shoot vertically, we experimented with camera cropping and UI treatments that leave enough space around the dish for a menu-perfect result.

With encouraging content and intentional visuals, we guide users in the moment while machine learning validates behind the scenes, creating a seamless flow that scales. Since launch, we’ve seen clearer, more useful photos and faster moderation. These improvements make food discovery on Uber Eats more reliable, helpful, and inspiring.