Hi, I’m Caleb. I shape ideas into products, and build the AI agents that power them.
Designing a Social Layer for MAX/HBO Max (Hackathon Winner)
Pitched and designed a social layer that turned content recommendations into virtual word-of-mouth — winning project at Warner Bros × BrainStation Hackathon.
The video below walks through the interactive Figma prototype.
Role
Problem
The hackathon brief: improve Max's content recommendation system to compete with other streaming platforms. The problem no solution addressed yet: algorithmic recommendations have a ceiling because they predict what you'll watch based on what you've watched. They can't replicate the single most powerful recommendation source that exists — another person whose taste you trust telling you to watch something.
My Scope
End-to-end UX/UI design
Concept origination and visual storytelling
Translating concepts to a team of 3 data science and 2 SWE
The Product
I proposed reframing the problem entirely: instead of building a better recommendation algorithm, build a social layer inside Max where users could see what people they cared about were actually watching.
Three core ideas:
Personal media profiles — users get a profile showing what they're watching now and what's on their watchlist, visible to others
Friend visibility — see what people you follow are currently watching, in real time
Celebrity Circles — opt into "circles" curated by celebrities, watching what your favorite stars watch
Together, these created what I called virtual word-of-mouth — recommendations that came from people, not predictions.
Below is a deeper look at the proposed social layer and "circles".
My Approach
Twenty-four hours with five teammates who didn't think visually meant the design had to do double duty: it had to convince judges, but first it had to convince my own team to bet on this concept over the safer "smarter algorithm" play.
I started with sketches to get the team aligned on the shape of the idea before any pixel work. Once we had buy-in, I moved fast through wireframes into the high-fidelity prototype.
Using the current MAX app as a skeleton, I was able to design a feature that looked seamless for mobile integration, getting the team back on track.
The MAX app was mixed with inspiration from successful features of other apps to design a mid-fi wireframe later conveyed to team. Hi-fidelity wireframes made post team-check in.
The Outcome
After a chaotic 24 hours, we pulled through with the win.
What shipped: The team won the Warner Bros × BrainStation Hackathon. Judges cited the social-layer concept as the most innovative submission and the prototype as the clearest articulation of a feature in the competition.
What I learned: Reframing the brief is a design move, not a violation of it. Most teams answered the literal question — "how do we improve content recommendations?" — with literal answers. The win came from treating the brief as a starting point and asking the deeper question underneath it: what's actually limiting recommendation quality, and is the answer technical or human?
