Strategy | UX Design | Development I Pitch Deck
Meex app
Year: 2026
Role: UX Strategist, UX Designer, Developer
Challenge: Social apps are designed to keep you scrolling. Meex was designed to make people put your phone away. That contradiction is the whole point, and also what made this hard to build and even harder to pitch. On top of the product problem, I did it alone: no team, no agency, no budget. Everything from strategy to prototype to investor deck had to come from me. The question I kept asking myself was how to make something that felt serious and credible without the infrastructure that usually produces serious, credible things.
Situation:
I kept noticing the same thing in university environments: students surrounded by people, sitting alone on their phones. Not because they didn't want to meet anyone, but because the gap between wanting to talk to someone and actually doing it had become impossible to cross. Apps didn't help. Most of them made it worse, replacing real awkwardness with digital avoidance.
Meex came from that observation. The idea was simple: show people who's nearby and willing to meet, give them something to say, and then get out of the way. No chat, no matching algorithm, no swiping, low rejection. The whole product is designed around a moment that happens in real life, not on a screen.
Task:
Because this was a solo project, there was no one to hand things off to. I had to move across disciplines that normally belong to different people entirely.
That meant researching and mapping the competitive landscape (twelve apps, assessed across how superficial they were, how algorithm-dependent, and how oriented toward actual in-person meetings), writing and stress-testing the business model, building the investor narrative, designing the product architecture, and producing a working interactive prototype, all while keeping a consistent point of view across all of it.
Action:
Strategy:
Working with Claude as a thinking partner, not a ghostwriter. I used Claude throughout this project, but not in the way most people use AI tools. I wasn't looking for it to write things for me, I was using it to pressure-test ideas, catch loose thinking, and move faster through the parts of the work that would otherwise have stalled me. When the revenue projections felt off, I had Claude check the arithmetic. When a slide felt like it was saying two things at once, I'd push on the framing until it was clean. The output always had to survive my own scrutiny, which meant Claude was most useful when I treated it like a smart collaborator I could argue with.
UX Design:
Building the deck programmatically Instead of building slides manually, I used pptxgenjs to generate the pitch deck through code. It sounds like extra work, but it meant I had full control over the content, structure, and data in a way that a regular presentation tool doesn't give you. When a number changed, everything updated. When a section needed restructuring, I could do it without rebuilding from scratch.
Where the AI lives inside the product The AI Opener is Meex's core differentiator, a feature that reads someone's profile and generates one specific, natural thing to say to them before you walk over. The design decision I'm most proud of is where it sits: before the approach, not during the conversation. Putting AI inside a conversation would have defeated the whole purpose of the app. Keeping it in the pre-approach moment means it builds confidence and then disappears. I designed that screen in the prototype so investors could experience the logic, not just read about it.
The prototype The full interactive prototype runs across eleven screens and covers the complete user journey, from toggling available, to browsing the map, to viewing a profile, to getting an AI-generated opener, to the moment the phone goes in the pocket. It was built in HTML and CSS. Using AI-assisted development meant the gap between having a clear idea of what a screen should do and actually seeing it work was much shorter than it would have been otherwise.
Result:
What came out the other side is a complete package: a 17-slide investor deck, a working prototype, a competitive analysis, a four-stream revenue model with verified projections, and a brand identity, produced by one person, in a timeline that would have been impossible without AI tooling used well.
The more interesting outcome for me is what I learned about how to work this way. AI didn't make this easier by doing the work. It made it faster by removing the friction between thinking something and testing it. The judgment still had to be mine. The decisions still had to hold up. But the gap between idea and execution got a lot smaller.
University pilot outreach is the current focus. The next version of this project is a signed agreement and a real campus — which is when it stops being a portfolio piece and starts being a company.
