CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Project Overview
Volta.AI is a Brazilian AI-powered digital loyalty platform that converts paper stamp cards to Apple Wallet/Google Pay, combined with intelligent marketing automation via WhatsApp campaigns. The platform targets Brazilian restaurants and high-end establishments, providing automated customer insights and campaign creation.
Key Features
- Digital wallet loyalty cards (Apple Wallet/Google Pay)
- QR code customer enrollment
- AI-powered campaign suggestions using DeepSeek API
- WhatsApp Business API integration
- LGPD-compliant customer data management
- Brazilian market-specific features (PIX, Portuguese localization)
Tech Stack
- Frontend: Next.js 15 with App Router, TypeScript, Tailwind CSS, shadcn/ui
- Backend: Supabase (PostgreSQL, Auth, Storage, Edge Functions)
- AI: DeepSeek API for text generation, FLUX.1 via fal.ai for image generation
- Deployment: Vercel
- Repository: https://github.com/pedrosantoshc/volta-ai.git
Development Commands
# Install dependencies
npm install
# Run development server
npm run dev
# Build for production
npm run build
# Type checking
npx tsc --noEmit
# Linting
npm run lint
# Add shadcn/ui components
npx shadcn@latest add [component-name]
Project Structure
src/
├── app/ # Next.js 15 App Router
│ ├── globals.css # Tailwind CSS styles
│ ├── layout.tsx # Root layout
│ └── page.tsx # Home page
├── components/
│ └── ui/ # shadcn/ui components
│ ├── button.tsx
│ ├── card.tsx
│ ├── input.tsx
│ └── label.tsx
└── lib/
└── utils.ts # Utility functions
Architecture Notes
- Mobile-first responsive design with progressive enhancement
- AI-first UX with prominent suggestions and one-click actions
- Row Level Security (RLS) for multi-tenant data isolation
- Brazilian localization throughout (Portuguese language, local business hours)
- LGPD compliance built into data models and user flows
Color Palette (from PRD)
- Primary: Purple gradient (#7c3aed to #a855f7) - AI/premium feel
- Success: Green (#16a34a) - positive metrics
- Warning: Amber (#d97706) - attention needed
- Background: Light gray (#f8fafc)
Current Development Phase
Phase 1B: Core Business Setup (MAJOR PROGRESS MADE)
Completed Tasks
✅ Next.js 15 project setup with TypeScript
✅ Tailwind CSS and shadcn/ui configuration
✅ Supabase integration (database, auth, storage)
✅ Authentication system (login/signup with Google OAuth)
✅ Basic dashboard layout with navigation
✅ Loyalty card designer with real-time preview
✅ Database schema deployed with all tables
✅ API documentation (PassKit, DeepSeek, FAL.AI)
✅ Development rules and workflow established
✅ Environment variables configured with API keys
Phase 1B Progress - Customer Management & LGPD Compliance
✅ Customer Management System - Full CRUD operations with advanced features ✅ Excel Import System - Dynamic template generation with loyalty card validation ✅ LGPD Compliance Framework - Complete consent tracking and validation ✅ Bulk Operations - Customer selection, bulk delete, CSV import/export ✅ Customer Segmentation - Active/inactive filters, search functionality ✅ Conditional Logic - Stamp validation against card limits, auto-completion ✅ Action System - Conditional button enabling based on consent status
Complete Development Roadmap
Phase 1B: Core Business Setup (MAJOR PROGRESS MADE)
1. Business Onboarding Flow ⏳ IN PROGRESS
Goal: Complete restaurant setup from signup to operational loyalty program
Components to Build:
/onboarding/step-1- Enhanced business registration (name, address, CNPJ, owner details)/onboarding/step-2- Logo upload to Supabase Storage + business type selection/onboarding/step-3- AI settings (tone: amigável/profissional/casual, brand voice, colors)/onboarding/step-4- Team management (invite staff, assign roles: owner/manager/staff)/onboarding/complete- Progress summary + first loyalty card creation
What This Accomplishes:
- Gets restaurants fully operational in under 10 minutes
- Sets up AI personality for Brazilian Portuguese campaigns
- Enables team collaboration with role-based access
- Ensures LGPD compliance from the start
2. Customer Enrollment System (MVP Heart) ⏳ IN PROGRESS
Goal: Convert walk-in customers to digital loyalty program members
Components to Build:
/enroll/[cardId]- Mobile-optimized public enrollment for QR scans/dashboard/cartoes/[id]/formulario- Dynamic form builder for custom questions- ✅ LGPD consent collection system (data processing, WhatsApp marketing, email) - COMPLETED
- ✅ Customer data processing pipeline with validation and deduplication - COMPLETED
- Automatic digital wallet pass generation and distribution
What This Accomplishes:
- Achieves 60%+ customer enrollment rate (PRD target)
- ✅ Collects valuable customer data for AI insights - COMPLETED
- Enables all downstream marketing automation
- Creates seamless mobile-first customer experience
Phase 1C: Core Loyalty Operations
3. Customer Management System ✅ COMPLETED
Components Built:
- ✅
/dashboard/clientes- Customer database with advanced search/filtering - COMPLETED - ✅ Customer segmentation builder (Active, Clientes eventuais, search filters) - COMPLETED
- ⏳ Individual customer profile pages with complete interaction history - PENDING
- ✅ Bulk operations (Excel import/export, bulk delete, customer selection) - COMPLETED
- ✅ LGPD-compliant consent tracking and validation - COMPLETED
- ✅ Conditional action system (buttons enabled based on consent) - COMPLETED
4. Manual Stamp Attribution System
Components to Build:
/dashboard/selos/adicionar- Quick stamp addition interface for staff/dashboard/selos/escaneamento- QR code scanner for customer identification- Bulk stamp operations for busy periods (add stamps to multiple customers)
- Complete audit trail (who added stamps, when, transaction history)
- Reward redemption tracking and notification system
Phase 1D: AI-Powered Insights
5. Basic AI Copilot Integration (DeepSeek API)
Components to Build:
- AI insights dashboard showing customer behavior patterns
- Automated campaign suggestions: "47 clientes inativos - Campanha 'Sentimos sua falta'"
- Expected engagement predictions and revenue impact estimates
- One-click campaign creation from AI suggestions
- Performance tracking for AI-generated campaigns
Key AI Features:
- Inactive customer detection (15+ days since last visit)
- Frequent visitor identification and VIP upgrade suggestions
- Birthday campaign automation
- Completed card reward reminders
6. WhatsApp Integration Foundation
Components to Build:
- WhatsApp Business API provider integration (360Dialog, Twilio, or ChatAPI)
- Template message system for Brazilian compliance
- Automated notifications (stamp additions, rewards available, card completed)
- Manual campaign sending to customer segments
- Message delivery tracking and analytics dashboard
Compliance Features:
- 24-hour messaging window compliance
- Template message approval workflow
- Unsubscribe mechanisms in all campaigns
- Rate limiting to prevent spam
Phase 2A: Digital Wallet Integration
7. Apple Wallet PassKit Implementation
Components to Build:
- PassKit certificate setup and management
- Dynamic pass generation with loyalty card data
- Real-time pass updates when stamps are added
- Pass distribution via enrollment flow
- Push notifications for pass changes
Technical Requirements:
- Apple Developer account and certificates
- PassKit Node.js SDK integration
- Pass signing and security implementation
8. Google Pay Integration
Components to Build:
- Google Pay API setup and authentication
- Pass generation for Android users
- Cross-platform wallet compatibility testing
- Unified pass management system
Phase 2B: Advanced AI Marketing
9. Advanced Campaign Generation
Components to Build:
- Natural language campaign generation with brand voice matching
- Customer segmentation AI based on behavior patterns
- A/B testing suggestions for campaign optimization
- Dynamic content personalization per customer segment
- Campaign performance prediction models
10. Visual Content Generation (FAL.AI FLUX)
Components to Build:
- Promotional image generation for WhatsApp campaigns
- AI-powered loyalty card background design
- Seasonal campaign imagery (Festa Junina, Natal, Carnaval)
- Brand-consistent visual asset creation
- Social media ready content in multiple formats
11. Advanced Analytics & ROI Tracking
Components to Build:
- Revenue attribution modeling linking campaigns to visits
- Customer lifetime value calculations
- Loyalty program ROI measurement dashboard
- Predictive analytics for customer churn
- Comparative analysis (loyalty vs non-loyalty customers)
Phase 3: Enterprise & Scaling
12. POS System Integration
Components to Build:
- Brazilian payment processor APIs (Stone, Cielo, GetNet)
- Real-time purchase tracking linked to loyalty cards
- Automatic stamp attribution based on spend thresholds
- Revenue correlation analysis and reporting
13. Advanced Enterprise Features
Components to Build:
- Multi-location support for restaurant chains
- Advanced role permissions and user hierarchies
- API for third-party integrations
- White-label solutions for payment processors
- Coalition loyalty programs (cross-brand partnerships)
Success Metrics (PRD Targets)
- Customer Acquisition: 60%+ enrollment rate at participating venues
- Engagement: 25%+ WhatsApp campaign engagement rates
- Performance: 99.5%+ uptime, <200ms dashboard response times
- ROI: 15:1 return on AI-generated campaigns
- Business Growth: R$ 100K+ ARR within 12 months
IMMEDIATE NEXT PRIORITIES (Current Focus)
High Priority - Core MVP Completion
-
Customer Enrollment Flow - Complete the public-facing customer registration
/enroll/[cardId]- Mobile-optimized enrollment page for QR scans- Dynamic form builder for custom business questions
- Integration with existing LGPD consent system
-
Individual Customer Actions - Replace placeholder alerts with real functionality
- Customer detail pages (
/dashboard/clientes/[id]) - Manual stamp addition interface
- Loyalty card status management
- Customer detail pages (
-
Digital Wallet Integration - Core value proposition
- Apple Wallet PassKit implementation
- Google Pay Pass generation and distribution
- Real-time pass updates when stamps are added
Medium Priority - Business Operations
-
Enhanced Onboarding Flow - Complete business setup experience
- Multi-step onboarding with logo upload
- AI personality configuration for campaigns
- Team member invitation and role management
-
Manual Stamp Attribution System - Staff operational tools
- QR code scanner for customer identification
- Bulk stamp operations for busy periods
- Complete audit trail and transaction history
Future Phases - AI & Marketing
-
Basic AI Copilot Integration - DeepSeek API
- Customer behavior analysis and insights
- Simple campaign suggestions based on patterns
- One-click campaign creation
-
WhatsApp Business API Integration - Customer communication
- Provider evaluation and selection
- Automated notifications for stamps/rewards
- Manual campaign sending to segments
Technical Priorities
- Mobile-first: All features must work perfectly on mobile devices
- Brazilian compliance: LGPD, Portuguese language, local business practices
- AI integration: DeepSeek for campaigns, FAL.AI for imagery
- Performance: <200ms dashboard loads, <500ms API responses
- Scalability: Multi-tenant architecture with Row Level Security
Development Approach
Following iterative development with UI/UX integrated throughout each phase rather than as a separate final step. WhatsApp integration planned as final implementation after API provider research.
STRATEGIC DEVELOPMENT PLAN - DECEMBER 2024
GPT vs Current Progress Comparison
After completing Phase 1B customer management and loyalty operations, here's how our current progress compares to the original GPT roadmap:
✅ MAJOR ACCOMPLISHMENTS (Ahead of Schedule)
Customer Management System - FULLY COMPLETED
- ✅ Advanced customer database with search, filtering, and bulk operations
- ✅ Excel import/export with dynamic template generation
- ✅ LGPD compliance framework with comprehensive consent tracking
- ✅ Customer segmentation (Active vs "Clientes eventuais")
- ✅ Conditional action system based on consent status
- ✅ Bulk delete operations with customer selection
- Status: GPT plan estimated this as Phase 1C, we completed in Phase 1B
Loyalty Card Operations - CORE FUNCTIONALITY COMPLETED
- ✅ Business scoping utility (
getCurrentBusinessId) - Fixed critical multi-tenant bug - ✅ Customer detail pages with loyalty card progress visualization
- ✅ "Dar selo" API (
/api/stamps/add) with business validation and auto-completion - ✅ GiveStampDialog component for intuitive stamp attribution
- ✅ Manual stamp attribution pages (
/dashboard/selos/adicionar) - ✅ Wallet integration stubs ready for Phase 2 implementation
- Status: GPT plan had this spread across Phases 1B-1C, we consolidated effectively
Technical Foundation - ROBUST AND SCALABLE
- ✅ Next.js 15 with proper TypeScript configuration
- ✅ Supabase integration with Row Level Security
- ✅ Mobile-first responsive design
- ✅ Error handling and user feedback systems
- ✅ API route structure following RESTful patterns
- Status: Exceeded GPT plan expectations for technical quality
🎯 CURRENT FOCUS AREAS (Next 4-6 Weeks)
1. Customer Enrollment System - HIGH PRIORITY
- ✅ Public enrollment page (
/enroll/[cardId]) - COMPLETED - ✅ LGPD consent collection with comprehensive compliance - COMPLETED
- ✅ Custom form builder integration - COMPLETED
- ⏳ QR code scanning functionality with camera integration - PENDING
- ⏳ Real-time QR code generation and management - PENDING
2. Digital Wallet Integration - CORE VALUE PROPOSITION
- ⏳ Apple Wallet PassKit implementation with certificate setup
- ⏳ Google Pay Pass generation and distribution
- ⏳ Real-time pass updates when stamps are added
- ⏳ Pass signing and security implementation
- Business Impact: This is the primary differentiator vs competitors
3. QR Code & Scanning System - OPERATIONAL EFFICIENCY
- ✅ Manual QR input functionality - COMPLETED
- ⏳ Camera integration for live QR scanning - PENDING
- ⏳ Customer identification and validation workflow - PENDING
- ✅ QR code scanning page structure - COMPLETED
📋 REFINED NEXT STEPS (Priority Order)
Phase 1C Completion (January 2025)
-
QR Scanning Implementation - IMMEDIATE PRIORITY
- Complete camera integration for live QR code scanning
- Implement QR code generation with customer card linking
- Add QR validation and error handling
- Files:
src/app/dashboard/selos/escaneamento/page.tsx, QR generation utilities
-
Digital Wallet Integration - CORE MVP FEATURE
- Apple Wallet PassKit certificate configuration
- Pass generation API endpoints (
/api/wallet/pass/[customerCardId]) - Google Pay Pass implementation
- Real-time pass updates via webhooks
- Files:
src/lib/wallet.ts, Apple/Google Pay API integration
-
Onboarding Flow Polish - USER EXPERIENCE
- Multi-step business setup with progress tracking
- Logo upload to Supabase Storage integration
- AI personality configuration (tone, brand voice)
- Team member invitation system with role management
Phase 2A (February-March 2025) - AI Integration
-
Basic AI Copilot - COMPETITIVE ADVANTAGE
- DeepSeek API integration for customer insights
- Inactive customer detection (15+ days since last visit)
- Simple campaign suggestions based on behavior patterns
- One-click campaign creation interface
- Files:
src/lib/ai-insights.ts, dashboard components
-
WhatsApp Business API - CUSTOMER COMMUNICATION
- Provider evaluation and selection (360Dialog vs Twilio vs ChatAPI)
- Template message system for Brazilian compliance
- Automated notifications for stamps/rewards
- Manual campaign sending to customer segments
- Message delivery tracking and analytics
Phase 2B (April-May 2025) - Advanced Features
-
Advanced AI Marketing - DIFFERENTIATION
- FLUX.1 integration via fal.ai for promotional imagery
- Natural language campaign generation with brand voice
- A/B testing suggestions and optimization
- Performance prediction models
-
POS Integration Foundation - ROI TRACKING
- Brazilian payment processor API research (Stone, Cielo, GetNet)
- Real-time purchase tracking linked to loyalty cards
- Revenue attribution modeling for campaign ROI
- Baseline vs performance comparison dashboard
🏗️ ARCHITECTURAL DECISIONS
What We Got Right:
- Business multi-tenancy with proper scoping prevents data leakage
- Component reusability (GiveStampDialog used across multiple pages)
- API design follows consistent patterns and error handling
- TypeScript interfaces properly model business logic
- Mobile-first approach ensures broad compatibility
Technical Debt to Address:
- Complete TypeScript
anytype elimination (currently in progress) - Implement comprehensive RLS policy audit
- Add unit tests for critical business logic
- Optimize database queries for larger datasets
- Add proper logging and monitoring integration
🎯 SUCCESS METRICS TRACKING
Completed Milestones:
- ✅ Customer management system supports 1000+ customers per business
- ✅ LGPD compliance framework prevents legal issues
- ✅ Manual stamp attribution enables operational efficiency
- ✅ Customer segmentation allows targeted marketing
Next Milestones:
- 📱 60%+ customer enrollment rate via QR scanning
- 💳 Apple Wallet/Google Pay integration working end-to-end
- 🤖 AI insights provide actionable customer behavior analysis
- 📈 WhatsApp campaigns achieve 25%+ engagement rates
🔄 STRATEGIC PIVOTS FROM ORIGINAL PLAN
What Changed:
- Customer Management Priority: Moved from Phase 1C to 1B due to operational importance
- LGPD Compliance Depth: Implemented more comprehensive framework than originally scoped
- Business Scoping: Added robust multi-tenant architecture not in original plan
- Component Architecture: Created more reusable components for operational efficiency
Why These Changes Worked:
- Earlier customer management completion enables better business validation
- Comprehensive LGPD compliance prevents future legal blockers
- Robust architecture supports scaling without major refactoring
- Operational tools (stamp attribution) provide immediate business value
💡 KEY INSIGHTS FOR NEXT PHASE
- QR Integration is Critical: Digital wallet value depends on seamless QR scanning
- AI Must Be Practical: Focus on actionable insights vs complex analytics
- WhatsApp Provider Research: Brazilian compliance requirements are complex
- Mobile Experience: All features must work perfectly on mobile devices
- Business Validation: Each feature should solve real restaurant operational pain points
This strategic plan positions us ahead of the original GPT timeline while maintaining focus on core business value delivery.
Development Rules & Documentation
IMPORTANT: Before implementing any feature, always follow these steps:
- Check PRD.md - Understand feature requirements and business context
- Review API Documentation - Check
/docs/folder for relevant API docs:passkit-api.md- Apple Wallet & Google Pay integrationdeepseek-api.md- AI campaign generation and insightsfal-ai-api.md- FLUX.1 image generation for marketingdevelopment-rules.md- Complete development workflow and standards
- Ask Questions - When information is missing or unclear, ask before proceeding
API Keys Configured
- ✅ DeepSeek: AI campaign generation and customer insights
- ✅ FAL.AI: FLUX.1 image generation for promotional content
- ✅ Supabase: Database, auth, and storage backend
- ⏳ PassKit: Digital wallet integration (to be configured)
- ⏳ WhatsApp Business API: Customer messaging (provider TBD)
Development Rules & Guidelines
Core Development Workflow
- Always check PRD first (
docs/PRD.md) for requirements and specifications - Check API documentation (
docs/passkit-api.md,docs/deepseek-api.md,docs/fal-ai-api.md) - Ask for clarification if information is missing or unclear
- Never assume API capabilities without checking documentation
Available API Documentation
- PassKit API (
docs/passkit-api.md) - Apple Wallet + Google Pay loyalty cards - DeepSeek API (
docs/deepseek-api.md) - AI campaign generation and customer insights - FAL.AI API (
docs/fal-ai-api.md) - FLUX image generation for promotional content - Development Rules (
docs/development-rules.md) - Complete development guidelines
API Keys Configured
- ✅ DeepSeek AI:
sk-a1ebdaeadc6b44459940d3bc40fb6f36 - ✅ FAL.AI FLUX:
791c05e6-beb4-4114-8036-64a52f39c6fa:4f5b9a7f03021af62598985027156208 - ✅ Supabase: Real production credentials configured
Important Notes
- Always maintain mobile-first approach
- Implement Portuguese localization from the start
- Ensure LGPD compliance in all data handling
- Use shadcn/ui components for consistent design
- Follow the established color palette for brand consistency
- ALWAYS check PRD and API documentation before implementing features
- Never assume API capabilities - Always check documentation first
- Test integrations thoroughly - Verify all API calls work correctly
Related Documents
Claude AI Git Workflow Integration
When working with git repositories and suggesting commit workflows, always recommend using the `git-ai-commit` CLI tool for creating commit messages.
Code indexing for AI agents: summarization strategies and evaluation systems
**The most effective code indexing systems combine hierarchical LLM-generated summaries with AST structural data and vector embeddings through hybrid retrieval—achieving up to 80% codebase reduction while maintaining high accuracy for AI coding agents.** Leading tools like Cursor, Sourcegraph Cody, and Continue.dev demonstrate that no single retrieval method suffices; production systems require semantic search, keyword matching, and structural queries working together. For evaluation, the field
Missing Business Agents Research — FLUXION 2026
> Deep Research CoVe 2026 | Date: 2026-03-23
write-script
Write a full video script for @SketchySurvival101 following all rules in CLAUDE.md.