Bountea UGC

A UGC marketplace where brands post bounties and creators win payouts with the help of AI validation.

Comment

Inspiration

The creator economy has exploded, but monetization remains opaque. Traditional brand partnerships are difficult to access, payment terms unclear, and validation processes manual and slow. We saw an opportunity to build a transparent, AI-powered platform that connects brands directly with creators while ensuring quality through intelligent validation. The name "Bountea" reflects our mission: brewing the perfect blend of opportunities for creators while making UGC campaigns as simple as having a cup of tea.

What it does

Bountea UGC is a comprehensive bounty platform that revolutionizes user-generated content campaigns:

  • For Businesses: Create bounties with specific requirements, set pay rates (per 1k views), and let AI automatically validate submissions. Track campaign progress in real-time with detailed analytics and manage your brand presence with custom logos and company profiles.

  • For Creators: Discover bounties across YouTube, TikTok, and Instagram. Submit content with automatic metadata extraction and view count tracking. Get paid based on actual performance with transparent progress tracking. All submissions are validated by AI to ensure they meet exact requirements—no more waiting days for manual review.

Key Features:

  • 🤖 AI-Powered Validation: Google Gemini AI analyzes video content against bounty requirements with 100% exact matching
  • 📊 Real-Time View Tracking: Automatic syncing of view counts from YouTube Data API and TikTok APIs
  • 🎯 Multi-Platform Support: Submit content from YouTube, TikTok, and Instagram seamlessly
  • 💰 Performance-Based Payments: Transparent pay-per-view model with real-time progress tracking
  • 🎨 Beautiful UI: Modern, responsive design with dark/light theme support
  • 🔐 Secure Authentication: Clerk-powered auth with role-based access (business vs creator)

How we built it

Frontend:

  • Next.js 16 (App Router) with TypeScript for type-safe, server-side rendered components
  • Tailwind CSS with custom animations and gradient effects for a polished UI
  • Framer Motion for smooth animations and transitions
  • React 19 with server components for optimal performance

Backend & APIs:

  • Next.js API Routes for serverless backend logic
  • Supabase (PostgreSQL) for database and file storage
  • Clerk for authentication and user management with webhook integration

AI & Integrations:

  • Google Gemini 2.5 Flash with structured JSON output for video content validation
  • YouTube Data API v3 for extracting video metadata and view counts
  • Peekalink API for TikTok video data and view tracking
  • LinkPreview API for fetching link metadata across platforms

Key Architecture Decisions:

  • Service role key pattern for server-side Supabase operations (bypasses RLS)
  • UUID-based IDs throughout (vs numeric IDs) for better scalability
  • CamelCase to snake_case mapping for API request/response handling
  • Graceful degradation when platform APIs are unavailable
  • Structured AI validation with strict exact-matching requirements

Challenges we ran into

  1. AI Validation Accuracy: Getting Gemini to validate content with 100% exact matching (not "similar enough") required extensive prompt engineering. The model initially accepted approximations, so we had to build strict validation schemas and detailed instructions emphasizing exact requirements.

  2. Multi-Platform API Integration: Each platform (YouTube, TikTok, Instagram) has different APIs, rate limits, and data formats. Handling failures gracefully while maintaining a consistent user experience required careful error handling and fallback mechanisms.

  3. Real-Time View Tracking: View counts need to stay current, but platform APIs have rate limits. We implemented smart caching and update strategies to balance freshness with API quota management.

  4. Video URL Validation: Extracting video IDs from various URL formats (youtube.com/watch?v=, youtu.be/, tiktok.com/, etc.) while handling edge cases required robust URL parsing and platform detection logic.

  5. Database Schema Evolution: Adapting from the original numeric ID system to UUID-based IDs while maintaining compatibility with existing patterns required careful migration planning.

  6. Clerk-Supabase Sync: Keeping user profiles synchronized between Clerk authentication and Supabase database through webhooks while handling race conditions was tricky.

Accomplishments that we're proud of

✅ AI-Powered Validation: Successfully implemented strict content validation using Google Gemini that ensures submissions match bounty requirements with exact precision—no approximations accepted.

✅ Multi-Platform Support: Built a unified submission flow that automatically detects platform (YouTube/TikTok/Instagram), extracts metadata, and tracks views from different API sources seamlessly.

✅ Beautiful, Modern UI: Created an elegant landing page with custom animations, gradient effects, and dark/light theme support that feels premium and professional.

✅ Real-Time Progress Tracking: Implemented dynamic bounty progress calculation based on approved submissions' view counts with live updates.

✅ Role-Based Access Control: Clean separation between business and creator roles with appropriate permissions and UI flows for each user type.

✅ Production-Ready Architecture: Built with scalability in mind using Next.js App Router, TypeScript throughout, and proper error handling patterns.

What we learned

  • AI Structured Outputs: Deep dive into Gemini's structured output capabilities, learning how to design schemas that enforce strict validation criteria through prompt engineering rather than post-processing.

  • Platform API Differences: Each social media platform's API has unique quirks—YouTube requires video ID extraction, TikTok needs special handling for view counts, and Instagram has limited API access. Building abstraction layers was key.

  • UUID vs Numeric IDs: Migrating to UUID-based IDs improved our data model but required careful consideration of type safety and database relationships.

  • Next.js App Router: Mastering Server Components, Server Actions, and API Routes in the new App Router paradigm significantly improved our application's performance and developer experience.

  • Clerk Webhooks: Proper webhook handling is crucial for keeping user data synchronized, but requires careful error handling and idempotency checks.

  • Error Boundary Patterns: Building robust error handling that gracefully degrades (e.g., falling back to link previews if platform APIs fail) while maintaining a good user experience.

What's next for Bountea UGC

🚀 Payment Integration: Implement Stripe Connect for automated payouts to creators based on their earnings, with escrow system for businesses.

📱 Mobile App: Native iOS and Android apps for creators to submit content on-the-go and receive push notifications for new bounties.

📈 Advanced Analytics: Comprehensive dashboard for businesses showing ROI, engagement metrics, and creator performance analytics. Creator dashboards with earnings forecasts and submission history.

🔍 Enhanced Discovery: AI-powered bounty matching based on creator's content history, performance metrics, and preferences. Search and filter improvements.

💬 Communication Tools: Built-in messaging between businesses and creators for clarification on requirements, feedback, and collaboration.

🌐 More Platforms: Expand beyond YouTube, TikTok, and Instagram to include Twitch, Twitter/X, LinkedIn, and other emerging platforms.

🤖 Smarter AI: Fine-tune validation models on custom datasets, add sentiment analysis, and detect authentic vs. fake engagement.

⚡ Performance Optimizations: Implement Redis caching for frequently accessed bounties, optimize database queries, and add CDN for media assets.

🎯 Gamification: Creator levels, badges, and achievements to incentivize quality submissions and platform engagement.

Built With

Share this project:

Updates