magma-production

byEthan Cross

Build a production-ready AI SaaS called “CloseQuote” for small residential painting contractors. Core promise: “Turn more painting estimates into approved jobs and paid deposits.” Target users: Painting contractors and small painting companies with 1–15 employees. Build a modern, clean, professional SaaS web app with responsive desktop/mobile UI. CORE WORKFLOW: 1. Contractor signs up/logs in. 2. Creates their business profile. 3. Adds customers. 4. Creates a quote with customer, project description, amount, quote PDF/link, expiry date, and deposit amount/percentage. 5. Quote gets a public customer-facing URL. 6. Contractor selects a follow-up schedule. 7. The system automatically follows up with customers by email. 8. AI generates personalized follow-up messages based on quote details, customer name, project, timing, and previous conversation. 9. Customer can view the quote, approve, decline, or ask a question. 10. AI analyzes customer replies and classifies intent such as interested, price objection, needs information, wants to delay, or not interested. 11. AI can draft a suggested response for the contractor, but the contractor must approve before sending. 12. When a quote is approved, the customer can pay the required deposit through Stripe. 13. Contractor dashboard shows quote status, follow-ups, approvals, deposits, conversion rate, and recovered revenue. MAIN PAGES: - Landing page - Sign up / Login - Onboarding - Dashboard - Quotes - Create/Edit Quote - Quote Details - Customers - Customer Details - Messages - Follow-up Automations - Analytics - Business Settings - Team Settings - Billing - Public Customer Quote Page QUOTE STATUSES: Draft, Sent, Viewed, Follow-up, Approved, Deposit Pending, Deposit Paid, Declined, Expired, Lost. AI FEATURES: - AI follow-up email generation - AI reply/intent classification - AI suggested customer responses - AI quote PDF information extraction - AI next-best-action recommendations - Conversation summaries IMPORTANT AI RULE: AI should assist and recommend, not directly change critical business states or send sensitive messages without contractor approval. Normal application logic controls authentication, authorization, quote status, payments, scheduling, and database operations. TECH STACK: - Next.js + React + TypeScript - Tailwind CSS - shadcn/ui - Supabase PostgreSQL - Supabase Auth - Supabase Storage for quote PDFs - OpenAI API for AI features - Stripe for deposits and SaaS subscriptions - Resend for email - Trigger.dev or equivalent background jobs for scheduled follow-ups - Zod for validation DATABASE: Use a multi-tenant architecture with organizations. Include tables for: organizations, profiles, organization_members, customers, quotes, quote_files, follow_up_sequences, follow_up_steps, messages, ai_interactions, payments, subscriptions. SECURITY: Use authentication and authorization with organization-level data isolation. Users must only access data belonging to their organization. Never expose API keys or payment secrets to the frontend. Use secure server-side API calls and database row-level security where appropriate. MVP PRIORITY: Focus on: Customer → Quote → AI Follow-up → Customer Response → Approval → Deposit → Revenue tracking. Do NOT build initially: Scheduling, payroll, invoicing, crew management, full CRM, mobile app, WhatsApp, SMS, Jobber/ServiceTitan integrations, complex workflow builders, or an AI chatbot. DESIGN: Premium but simple B2B SaaS aesthetic. Make the dashboard immediately understandable. Emphasize “quotes needing attention,” “AI recommendations,” “approved revenue,” and “deposits collected.” Use cards, tables, status badges, clear CTAs, empty states, loading states, error states, and responsive layouts. Seed the application with realistic demo painting-business data so the dashboard looks useful immediately. Build the application with clean, modular, maintainable code and production-ready error handling. Prioritize a working end-to-end MVP over adding unnecessary features.