Vibeathon Winners! S02

Meet the winners of Vibeathon S02. Builders who pushed the limits of design, code, film, marketing and gaming to create what’s next.

Vibeathon Winners! S02

Over the past 4 weeks, builders from across the B150 and the world, from over 14 countries, pushed their creativity to new levels. From interactive games to generative films to new AI-powered tools, this Vibeathon S02 showcased what’s possible when you combine imagination with the right technology.

Participants were judged based on the following criteria:

Final scores is a combination of Judge Scores (70%) and Community Score(30%).

1. Potential Impact
• Does it solve a meaningful problem or open up new creative possibilities?
• Could it reach or inspire a lot of people?
• How viral, useful, or emotionally resonant is it?

2. Originality & Creativity
• Is the concept fresh, bold, or inventive?
• Does it stand out from what’s already out there?
• Does it feel like something only this person/team could have made?

3. Execution & Technical Craft
• Is the build functional, polished, and technically sound?
• Does it make clever use of AI tools or workflows?
• How thoughtful is the structure, flow, and depth of the project?

4. Design & User Experience
• Is the design clean, intentional, and accessible?
• Does the experience feel cohesive, delightful, or intuitive?
• How well does it balance form and function?

Thanks to our amazing Sponsors:

Thanks to our judges: Adam Perlis Martha B. Melissa Diago Linus Ekenstam Marten Kuipers Emmet Connoly Gizem Akdag Saket Ozarka Caroline Ciaramitaro Siya Mali Elsa Welshofer Robyn Park Boris Starkov Paul Faivret Elizabeth Lin

Today, we’re excited to announce the winners...

Vibe Coding

Winner:

Sujit Meka, United States

Project Name:

Hollistic HealthAI

Description:

"OpenHealth - Your Holistic AI Health Co-Pilot

Spent 4 years figuring out health through trial and error. Built this so you don't have to.

What it does:

Combines body scans + bloodwork + daily activity (Whoop, Apple Watch, etc.)

AI sees the full picture - not just isolated metrics

Co-creates diet, exercise, and supplement plans with you

All backed by credible sources (PubMed, peer-reviewed research)

The problem: Health advice is fragmented and often wrong. Fitness bro tells you to slam 2 Raising Canes buckets for protein. Bloodwork shows high LDL, doctor says cut oil.

Reality? Maybe you need more soluble fiber. Or HIIT workouts. Or sauna sessions. Or cacao powder. All of these help - but to what efficacy? In what combination? Nobody knows.

Macros aren't the full picture. True health is interconnected - compounds, nutrients, therapies all working together. It's overwhelming to understand, let alone optimize.

OpenHealth changes this: The agent sees it all. Creates plans that translate to real outcomes. Significantly reduces the barrier to personalized health.

Try it free - your data stays local, AI calls use your APIs. You're in control."

Link: https://github.com/sujitmeka/OpenHealth/tree/main

Prize:

1-year subscription to v0 + Base44 + Kiro


Runner-Up:

Dan Posalski, Israel

Project Name:

Creyone

Description:

"Everything started to help my sister and then my wife envision a renovation project.

I've built an app that creates images from 3d sketch, from prompt - but with the ability to change any element into a shoppable item

Moving from inspiration to shopping."

Link: https://creyone.com/

Prize:

1-month subscription to v0 + Base44 + Kiro

Vibe Gaming

Winner:

Sophia Zhang, United States

Project Name:

Rat in a Fridge Simulator

Description:

What if your pet got trapped somewhere and you didn't notice?

That anxious thought became this rhythm-typing game. You play as Noodle, a diabetic rat accidentally locked in a fridge, typing letters to keep him conscious enough to signal his owner Maya before time runs out.

The words you type trace his journey: from "COLD, NUMB, WEAK" to "LIGHT, SAFE, HOME"—if you can keep him alive. Every keystroke is a heartbeat. The rhythm IS the story.

10-15 minutes. One rat. One chance.

My friend Ryan and I built this with Phaser 3, vibe-coded with Replit and Cursor, and fueled by intrusive thoughts about pet safety.

Go hug your pets after playing.

Link: https://so2zhang.itch.io/rat-in-fridge-simulator

Prize:

Winner: 1-year subscription to Replit Core


Runner-Up:

Guillermo Gonzalez, Mexico

Project Name:

🤓 Nerdy word guessing game 🤔

Description:

While learning how LLMs work, I became fascinated by their ability to map semantic similarity between concepts.

To explore this, I created a game that uses real embeddings. In this game, players must guess a word that matches a specific target semantic similarity.

In non-technical terms: you must guess a word that is semantically identical, opposite, or unrelated to a given word in just the right amount to dominate the scoreboard.

Fully built in Replit, with a modern tech UI and a sneak peek into the magic in LLMs

🕹 See you in the scoreboard! 👊

Link: https://semantic-explorer.replit.app/

Prize:

Runner-up: 1-month subscription to Replit Core

Vibe Music/Audio

Winner:

Linda Liu, United States

Project Name:

Boston Pulse: A Daily Listen

Description:

Boston Pulse grew out of a personal gap I noticed as a BU student and immigrant studying communications and data science.

I’ve always known that staying informed matters, especially as a communications student and as part of a broader community, but I struggled to find a way of engaging with news that felt sustainable for me. Traditional feeds often felt overwhelming, and constant screen-based consumption pushed me further away rather than helping me stay connected. I wanted a way that felt less stressful and more accessible to keep up with what’s happening around me.

Boston Pulse curates three real editorial excerpts each day from Boston-area outlets and turns them into a calm, audio-first listening experience for staying informed on the go—whether you’re brushing your teeth, going through a morning routine, or riding the MBTA—without staring at your phone. The stories are selected from trusted public RSS feeds, with an emphasis on credible local reporting and editorial substance. Instead of AI-written summaries, it reads short passages directly from the original articles, preserving the writer’s voice and valuing the editorial work behind each piece, with links to the full story if you want to explore further.

Built as a rapid vibe-coding experiment, Boston Pulse explores how AI can support listening through a simple, minimal experience that respects journalism and reduces screen fatigue.

🎧Three stories. Boston, as it is today.

Link: https://boston-pulse-vibe.vercel.app/

Prize:

Winner: 6-month subscription to ElevenLabs Scale


Runner-Up:

Ersin Koç, Türkiye

Project Name:

I am the World - My first AI music video

Description:

Hello everybody. I am very excited to share my very first AI music video with the world. I'd like to hear your thoughts and suggestions about it.

Ben Dünya'yım / I am the World" is a poem I wrote some time ago. It was the end of a cycle and a beginning of a new one for me and was emotionally intense; a re-collection of years of experience and waking of awareness starting to have real form. This year I came up with the idea to turn those words into an audio-visual experience by using the latest AI tools.

As I have been using these tools for a while for my professional work, I am now able to build these worlds with a story; I craft consistent visual styles, make music for the mood and flow of the story, generate and edit visuals with intent and consistency. Then I turn them into cohesive videos with clear direction and emotional continuity, shaping them as a single experience rather than a sequence of scenes.

I used AI tools also for editorial purposes within this project. For example, I’ve developed the following paragraphs with AI assistance.

The poem “Ben Dünya’yım / I am the World” speaks from a place where identity dissolves and reforms continuously—where the self is not fixed, but fluid; not singular, but multiple; not separate from the world, but identical to it. The voice moves freely between “I,” “you,” “we,” and “one,” mirroring a state of awareness in which all boundaries soften.

The audiovisual work follows this same rhythm. Images shift, repeat, and transform, echoing the poem’s cycles of becoming, dissolving, and returning. Sound and music do not accompany the words; they inhabit them. Together, they form a space where time feels circular and existence unfolds as a constant movement rather than a linear progression.

This project is not intended to explain a concept, but to evoke a state—one of remembering, multiplicity, and quiet unity. It reflects a moment when accumulated experience stops being personal history and becomes awareness itself.

“Ben Dünya’yım / I am the World” marks the end of one cycle and the beginning of another.

This song and video mark the beginning of a larger body of work. “Ben Dünya’yım” is envisioned as an album that will consist of multiple pieces, each drawing from different parts of the same poem. Future tracks will approach the text from distinct emotional and perceptual angles, allowing the words to gradually unfold through sound and image rather than being presented all at once.

Together, these works are intended to form a unified audiovisual cycle, in which recurring themes, moods, and visual languages echo one another. Each piece is designed to stand on its own, while also contributing to a broader whole—reflecting the poem’s central idea of multiplicity emerging from a single source.

Director’s Note on AI Authorship

AI tools were used throughout this project as part of the creative process, functioning as instruments rather than authors. All conceptual decisions—including the interpretation of the poem, the emotional direction, visual language, musical structure, pacing, and final editorial form—were made by me.

The poem, the overarching concept, and the narrative intent originate from my own authorship. AI systems were employed to generate and transform visual and sonic material under direct creative guidance, similar to how one might work with complex digital software or generative instruments. Each output was curated, refined, edited, and assembled with deliberate intention.

In this sense, AI serves as a medium through which ideas take shape, not as an autonomous creator. The work remains grounded in human authorship, memory, and awareness, using contemporary tools to explore timeless questions of identity, transformation, and unity.

Tools used in the making of this music video:

#ChatGPT for editorial assistance, theme&prompt exploration,

#Midjourney for developing visual style and image generation,

#NanoBanana for image editing,

#Kling for video generation,

#Freepik as the platform.

Prize:

Runner-up: 3-month subscription to ElevenLabs Pro

Vibe Filmmaking

Winner:

Kevin Shah, United States

Project Name:

The Secret Life of Anjali

Description:

In war-torn South Asia, young Anjali has lost her father. Night after night she begs her mother to retell his secret bedtime story: a cosmic lullaby of reincarnation, love, destruction and home that stretches across 4.5 billion years. As the mother’s voice rises, Anjali’s imagination erupts. 

The Secret Life of Anjali could only be fully realized through the current tech. AI unlocked a visual and sonic language previously impossible on an independent scale.  All imagery was created using Google Veo in Flow, ElevenLabs generated the child’s voice, sound effects, and reference audio. I used Suno to create the score.  Human collaborators provided the final Hindi voice performance (Akanksha Mishra), translation (Tanvi Appa), and editing and sound design happened in Adobe Premiere Pro. Grok and Claude assisted with subtitle proofreading.

My hybrid approach; AI as primary visual, sonic, and creative collaborator paired with deliberate human direction, performance, and curation — allowed me to conjure an entire poetic world while preserving human truth and cultural resonance. The AI here is not a shortcut; it is the brush that made this intimate and emotional cosmos possible.

I'm very excited to share it here with you for Vibeathon!

Prize:

Winner: 1-year subscription to Freepik Essential


Runner-Up:

Dave Spector, United States

Project Name:

AI Can't Act

Description:

The goal of this short is to demonstrate nuance in an AI performance. Eye rolls, hesitations, nose rubs, mid-sentence shifts in tone, and emotions. And pores.

All of the performances were generated in Veo 3 from prompts. Lots of experimentation to figure out how to be push AI's buttons to get the performances I wanted.

Prize:

Winner: 1-year subscription to Freepik Essential

Vibe Marketing

Winner:

Viktor Konovalov, Poland

Project Name:

[*] Pokpok.ai. The House MD of Market Truth

Description:

Your competitors know something you don't.


WE’RE CRAZY ENOUGH TO TRY TO MAKE OUR $5K ENTERPRISE REPORTS ACCESSIBLE TO INDIE FOUNDERS FOR A FRACTION OF THE PRICE. THAT MEANS WE HAVE TO SELL 100x MORE THAN BEFORE TO KEEP THIS THING GOING — COMPARED TO CORPORATE PRICING

EACH SALE IS A REAL, SLIGHTLY TERRIFYING PROOF THAT ALL THIS PAIN — AND THIS ALL-IN BET — IS WORTH IT.

CAN’T WAIT TO SEE WHAT YOU’LL DO WITH THE SAME ENTERPRISE-GRADE KNOWLEDGE.

JUST PROMISE YOU’LL BUILD INNOVATIVE PRODUCTS. CHANGE THE GAME. FIGHT BACK.

We analyze 100% of your category—every brand claim, every product page, every customer review—and diagnose the three truths your competitors can't see.

WORKFLOW: n8n (automation of personal 25+ years strategic branding experience as Creative Director), Perplexity (helping to find relevant product links), Gemini 3PRO, Supabase, Polar, Claude Code, Figma -> Framer, Google notebookLM - podcast, Freepik (mascot), ElevenLabs (voice cloning)

Still working on (coding by myself). F*g crazy challenge. Your feedback is very much needed!

Link: https://pokpok.ai/

Prize:

Winner: 1-year subscription Flodesk + N8N


Runner-up:

Kathy Wang, United States

Project Name:

Press Club

Description:

i run press club, and we help founders land press. my co-founder didn't want to pay $5k for a pr agency, so he locked in and spent a week sending 200 cold emails to journalists. it worked, and he landed forbes and the wall street journal (and 8 other outlets). we built an ai agent to automate that type of manual outreach 😎

now, that's all fine and dandy, but here's the problem. people kept getting confused as to what we actually did — were we a b2b saas ai agent? were we actually a tech-enabled service? were we another pr agency?? (plot spoiler: we're a tech-enabled service right now.) so alas, i realized it was time to update our landing page.

i built the entire landing page with lovable in a span of 2 hours, but that wasn't enough. i thought, why don't we include a lead capture form that allows people to assess their own reportability (how reportable am i??).

i built it, and that was fine and dandy, but i wanted to take it ONE step further: capture all the emails + linkedins and enter it into our CRM. that's when n8n came in.

and okay, okay...i know this has been done a million times (the lead capture form to CRM automation), but i'm a DESIGNER, okay? the last time i took a computer science class was 7 years ago.

so you can SEE why i'm so excited about this. this was my first time using a workflow automation like n8n and learning about webhooks and POST requests and notion integrations and it WORKED PERFECTLY.

whenever someone takes the cute lil reportablity quiz and enters their email + linkedin, my notion CRM updates with their email + linkedin. it sounds simple, but it's ALL I EVER WANTED 🥺

Link: https://pressclub.lovable.app/

Prize:

Runner-up: 1-month subscription Flodesk + N8N 

Vibe Community Award

Winner:

Dee Pa, L

Project Name:

Moments: A quiet app for capturing life as it happens

Description:

What if your day could turn into a video you actually want to rewatch?

Moments is an iOS app that makes it effortless to capture and relive your day.

Instead of recording one long video or forgetting to record at all, you simply capture multiple 5-second clips throughout the day - small moments, nothing staged.

At the end of the day, Moments lets you stitch those clips into a single video diary.
(Automatic end-of-day stitching is coming next.)

How it works

Record short 5-second clips anytime during the day

Moments collects them quietly in the background

Generate a stitched video diary of your day with one tap

Rewatch it, save it, or share it with others

It’s designed to feel light, fun, and natural, so you capture more moments without thinking about it.

Why I built it

A friend of mine was already doing video diaries, but the process was heavy -> recording clips, stitching them together, and managing everything manually.

I loved the idea, but wanted to remove the friction.

Moments keeps the joy of video diaries while making them easy and accessible for anyone.

I’ve attached a short recording of me using the app, showing how clips turn into a stitched video diary
(shoutout to my partner for being my human tripod 😂).

I’d love to hear your feedback:

Does it make you want to record more moments?

Does the stitched diary feel meaningful to watch or share?

Built during Vibeathon 🖤
— Dee

Link: https://testflight.apple.com/join/acKVj3s1

Prize:

Winner: 6 month subscription to Perplexity Pro

A Message to the Winners:

Our team will reach out to each of you directly to help coordinate the prize redemption. Stay tuned.

All winners will receive a 6 month subscription to Perplexity Pro.

What Comes Next?

These projects represent what the Builder Economy is all about: experimenting in public, learning together, and showing what’s possible when creativity meets AI.

Congratulations to all the winners and participants. You set the bar for the next Vibeathon S02. We can’t wait to see what you build next

Stay tuned for Vibeathon S03...