Portfolio /
We built Rizz AI as a one-tap AI photo generator — turning any selfie into a professionally styled portrait, chosen from a library of hundreds of templates, without a single prompt typed.
We shaped the experience around one idea: great AI photos shouldn't require knowing how to talk to AI. Pick a style, tap once, and the app handles the rest.
Services
Scroll to explore
Complexity Simplified
The harder problem sat underneath the surface: build hundreds of styles that still look like the person in them, not a stranger wearing their face, and make all of it reachable in one tap, with zero prompts and zero decision fatigue over which of hundreds of options to pick.
Strategy In Action
We started by mapping how people actually browse for a look - not a workflow, a mood. From there, we designed around a single tap: upload a photo, pick a style, and the generation happens entirely in the background while you keep scrolling.
Every template was built to hide the hard part, not show it off. Hundreds of styles sit inside categories people already think in - headshots, dating, travel, aesthetic - so choosing one feels like scrolling a feed, not filling out a brief. By the time you'd normally still be picking a filter, the finished photo is already sitting in your gallery.
Development
Rizz AI turns a selected style and a selfie into a portrait-generation journey. The development direction focuses on input quality, a clear generation state, and a comparison experience that keeps the original photo close to the result.
Organize templates by the occasion and outcome someone is looking for. Keep the input requirements attached to the style, including flows that need two people.
Guide camera and gallery selection with clear photo requirements. Show what makes a useful input before asking someone to commit to a generation.
Give generation requests explicit waiting, success, and retry states. Preserve the selected style and source photos so a failed attempt does not require starting again.
Keep original and generated images easy to compare, then make the next action clear. Review single-person and two-person templates as distinct journeys.
Technology stack
Proposed stackA technology direction for this product.
A shared foundation for the mobile experience.
Native device capabilities and app delivery.
Connected data and account services.
Typed interfaces across the application.
People don't come to Rizz AI wanting "a photo" - they want a Christmas card photo, a birthday post, a bridal shoot, or whatever meme format is trending that week. So the library isn't one big pile of templates; it's organised around the actual reasons someone opens the app, occasion by occasion, trend by trend - so there's rarely a moment where the thing you needed didn't exist yet.
The result never replaces the original outright - it sits right beside it, one swipe away, so you can drag between "you" and "you in this style" until it actually sinks in. No separate editing step, no waiting between generation and reveal. One tap got you here; one swipe is all it takes to trust it.
Every other template works from one face. This one is built for two — upload a photo of you, a photo of your partner, and both get placed into the same scene together, not stitched or pasted, but generated as one photo. It's the one style built for a moment you're never actually in alone.
A generated photo is only as good as the selfies it's built from — so instead of letting people upload anything and risk a disappointing result, the app shows real good and bad examples first: no sunglasses, no group shots, no heavy filters. One extra screen now saves a wasted generation and a bad first impression later.
Camera · Gallery
