VibeChexk
An AI-Powered Social Recognition App for Gen Z
My Role: Product Owner & Solo Builder
VibeChexk App Interface
Hero Image Placeholder
Overview
MY ROLE
Product Owner, Builder, Growth Strategist
TECH STACK
Lovable, Bolt.new, OpenAI API
TEAM SIZE
1 (Solo Project)
KEY RESULTS
Acquired 200+ users in the first week, Achieved a 15% week-1 retention rate, Executed a zero-budget viral launch.
The Problem: The "Positivity Gap" in Gen Z Social Media
While observing social media trends, I identified a clear "positivity gap" for Gen Z users. Existing platforms often feel like high-pressure environments focused on curated perfection. There was a clear market appetite for low-stakes, positive, and authentic social interactions, a need once filled by the popular but now-defunct app "TBH." The core problem was clear: could I build a modern, AI-enhanced successor that captured this desire for positive recognition among peers?
The Solution: A Low-Stakes App for Positive Social Recognition
My solution was VibeChexk, an app designed from the ground up to be a fun, low-stakes environment for positive social interaction. The core idea is simple: users send and receive anonymous positive "vibes" through fun, AI-generated polls about their friends. It's a modern successor to the popular app "TBH," but with an AI-powered engine designed to maximize engagement and retention.
How It Works: The Core User Loop
Receive a Vibe
The experience begins when a user receives a poll. The app's AI engine generates a positive, context-aware question like, "Who has the most contagious laugh?" and presents four friends as options. The user anonymously votes for one friend, sending them a positive "vibe."
Main Polling Screen
AI-Generated Question with Friend Options
Notification Screen
"You Received a Vibe!" Display
Send a Vibe
When a user is selected in a poll, they receive a notification that a friend has sent them a "vibe." They can see the compliment but not who sent it. This creates a positive feedback loop and a compelling reason to stay engaged with the app and their friend group.
Climb the Leaderboard
To drive long-term retention, I integrated gamification elements. Users earn points for sending and receiving vibes, which places them on a weekly leaderboard among their friends. Daily streaks and rewards were also implemented to create a compelling reason for users to return to the app every day.
Leaderboard & Daily Streaks
Gamification Features Display
The Build: Engineering the MVP as a Product Owner
1. The "No-Code First" Philosophy & Tech Stack
As a Product builder, my primary constraints were time and capital. My strategy was to adopt a "no-code first" philosophy to maximize development velocity and validate the product with real users as quickly as possible. I chose a modern, efficient stack: Lovable for the back-end logic and user management, and Bolt.new for building the front-end and interactive prototype. This approach allowed me to focus on the user experience and the core AI features rather than on boilerplate infrastructure.
Tech Stack Architecture
Lovable + Bolt.new + AI APIs Diagram
2. The AI Engine, Powered by Google Gemini
The heart of VibeChexk is its AI pipeline, which I designed and integrated to drive user retention. For this critical component, I relied heavily on Google's Gemini model as my primary AI co-pilot and engine. I used Gemini for virtually every aspect of the development process: from generating front-end components in Bolt to designing the database schema and writing the back-end logic for the AI pipeline itself. The core AI features included:
Dynamic Prompt Engineering
I used Gemini to develop a system that could generate context-aware, personalized polling questions based on user inputs, ensuring the content never felt repetitive.
Sentiment Analysis
I implemented a sentiment analysis model to filter and prioritize positive interactions, which was crucial for reinforcing the app's core value proposition of positivity.
API Integration
I leveraged Gemini to write and debug the code needed to seamlessly integrate these AI features with the Lovable back-end.
AI Pipeline Diagram
User Data → Gemini API → Personalized Output
Gemini Prompt Example
Code Snippet Visualization
3. Prototyping, Iteration, and Business Impact
Using Bolt, I built a fully interactive, high-fidelity prototype that I could put directly in the hands of potential users. This allowed for a rapid feedback loop, enabling me to iterate on the UI and the AI-driven features before the public launch. This hands-on, AI-assisted approach was not just a technical choice; it was a strategic business decision. By building the entire MVP myself with the help of Gemini, I was able to de-risk the project and effectively saved the significant early-stage cost of hiring a dedicated front-end engineer.
Interactive Prototype in Action
Mobile Device Mockup - High-Fidelity Prototype
The Launch Strategy: Engineering Virality from Day One
So, how do you launch a social app for a Gen Z audience with a zero-dollar marketing budget? You go where they live: Instagram. My strategy wasn't just to post; it was to create a campus-wide mystery, a story that people wanted to be a part of. Here's how we did it, phase by phase.
Phase 1: Building Hype & a Captive Audience
Our first move was to create a private Instagram account named after the app and our college. For one week, we did nothing but follow students—nearly every single person on campus. We deliberately didn't accept any follow requests back. This single action created a powerful sense of curiosity and FOMO. Who was behind this page? What was it for? By the end of the week, we had hundreds of pending requests, creating a captive audience that was eager to see what was next.
Private Instagram Profile
"Requested" Button Visible
Phase 2: The Reveal: Turning Curiosity into Action
Timing was everything. At a peak social time—Saturday morning, around 11 AM—we made our move. We accepted every single follow request simultaneously. The account's first and only post was a visually appealing graphic with a simple, direct call to action: "Link in Bio." The bio, of course, contained the direct link to download the VibeChexk app. The sudden reveal to our now-captive audience was designed to create an immediate, concentrated burst of downloads.
Instagram Post & Bio
"Link in Bio" CTA with Download Link
Phase 3: Fueling the Flywheel for Organic Growth
The launch was just the beginning. The real magic was in the product's core mechanic. The app was designed to be inherently viral. When a user sent an anonymous compliment or "vibe" to a friend, that friend would get a notification. To see who sent it, they had to download the app. This created a powerful, self-sustaining viral loop that turned our initial user base into an organic growth engine, all without spending a single dollar on ads.
Viral Loop Diagram
User A → Sends Vibe → User B Gets Notification →
Downloads App → Sends to User C
Impact & Results
The strategic launch was a success, proving that a well-designed growth strategy can be more effective than a large budget. The app achieved strong initial traction and validated the core product hypothesis.
200+
Users Acquired
In the first week of the pilot launch.
15%
Week-1 Retention
Demonstrating strong initial product-market fit.
Learnings & Future Steps
Building and launching VibeChexk solo was an invaluable lesson in end-to-end product ownership. My key learning was the power of combining a user-centric product with a creative, user-centric launch strategy. The next step for VibeChexk would be to analyze the initial user data to double down on the most engaging features and build a more robust V2 to expand beyond the initial college community.