• 

  Onboard3

• 

  Onboard2

• 

  User Coupons

• 

  User Profile

• 

  Onboard1

cultureQ Project Documentation

Inspiration

Generic coupon apps spam users with irrelevant offers, while banks sit on rich transaction data that could power truly personalized recommendations. We saw an opportunity to bridge this gap—using actual spending patterns to deliver coupons people actually want to use.

What it does

cultureQ analyzes your banking transactions to understand your taste preferences, then delivers personalized brand recommendations and cashback offers. The platform:

• Connects securely to your bank account via Plaid
• Maps transactions to brand preferences using Qloo's taste intelligence API
• Provides conversational AI to refine your taste profile
• Delivers targeted offers from partner merchants
• Tracks cashback and redemption analytics

How we built it

Frontend: Next.js 15 with React 19, TypeScript, and Tailwind CSS for a responsive user experience

Backend: Convex real-time database with serverless functions for scalable data processing

Integrations:

• Plaid API for secure banking connections
• Qloo API for AI-powered brand recommendations
• OpenAI for conversational taste profiling

Architecture: Context-driven state management with caching for performance, admin dashboard for merchant management, and RESTful APIs for external integrations.

Challenges we ran into

Transaction categorization: Mapping messy bank transaction data to meaningful brand preferences required extensive data cleaning and fuzzy matching algorithms.

Privacy concerns: Balancing personalization with user privacy—we anonymize transaction data while preserving recommendation accuracy.

API rate limits: Managing multiple external API calls (Plaid, Qloo, OpenAI) without hitting rate limits required intelligent caching and request batching.

Real-time updates: Synchronizing transaction data, taste profiles, and offers across the platform while maintaining performance.

Accomplishments that we're proud of

• Built a fully functional fintech application with secure banking integration
• Created an intelligent taste profiling system that learns from actual spending behavior
• Developed a scalable admin system for merchant onboarding and offer management
• Implemented real-time data synchronization across multiple APIs
• Achieved seamless user experience from account linking to coupon redemption

What we learned

Data is messy: Real-world transaction data requires significant preprocessing before it becomes useful for AI recommendations.

User trust is everything: In fintech, transparent privacy practices and clear value propositions are crucial for adoption.

API orchestration complexity: Coordinating multiple external services requires robust error handling and fallback strategies.

Personalization at scale: Building recommendation systems that work for diverse user bases while maintaining relevance is both an art and science.

What's next for cultureQ

Enhanced AI: Implement machine learning models trained on our transaction data for even better taste prediction accuracy.

Merchant expansion: Build partnerships with major retailers and local businesses to expand our offer ecosystem.

Social features: Allow users to share taste profiles and discover offers through their networks.

Advanced analytics: Provide merchants with detailed insights into customer preferences and campaign performance.

Mobile app: Develop native iOS/Android apps with location-based offer notifications and in-store redemption features.

Built With

• convex
• nextjs