Can I vibe code Zendesk?

Build a production-ready, multi-tenant Customer Support Desk and SLA Orchestration Engine designed for e-commerce merchants.

DATA MODELS & SCHEMA:
- Core entities: Tenant, User (roles: admin, agent, viewer), Customer, Ticket, ThreadMessage, Macro, KnowledgeArticle, SLA_Policy, TriggerRule, ChannelConfig.
- Ticket model fields: id, tenant_id, customer_id, subject, status (open, pending, solved, closed), priority (low, medium, high, urgent), channel (email, chat, api, shopify), assigned_agent_id, group_id, tags (array), custom_fields (JSONB), sla_due_at, first_responded_at, created_at, updated_at.
- ThreadMessage fields: id, ticket_id, sender_type (customer, agent, system, ai_agent), sender_id, body_html, body_text, attachments (JSONB array), is_internal_note (boolean), external_id (for email Message-ID tracking).

INGESTION & DELIVERY PIPELINES:
- Implement an inbound email parser handling Webhook payloads from SendGrid/Postmark. Extract headers (In-Reply-To, References, Message-ID) to thread emails correctly into existing tickets or construct new ones. Handle multi-part MIME parsing and S3 file upload attachments.
- Construct an outbound email sender using AWS SES/Postmark API. Automatically inject tracking headers (References, In-Reply-To) to ensure customer email clients maintain native thread views.
- Implement a WebSocket server for real-time live chat widget communication, handling agent presence typing indicators, read receipts, and session handoffs.

CORE TICKET ENGINE & WORKFLOWS:
- Build an Execution Engine for Triggers and Automations (e.g., IF status == open AND tag includes 'return' THEN assign_group = 'Returns' AND set_priority = 'high').
- Create a SLA tracking engine running on a background worker (Redis/BullMQ) calculating First Response Time (FRT) and Next Response Time (NRT) based on tenant-configured business hours schedules. Trigger webhook alerts on SLA breaches.
- Build an enterprise Macro Engine supporting dynamic Liquid tags (e.g., {{ customer.first_name }}, {{ ticket.id }}, {{ shopify.latest_order.tracking_number }}).

SHOPIFY & E-COMMERCE INTEGRATION:
- Build a side-panel data provider that queries Shopify GraphQL Admin API using the customer's email address. Render order history, fulfillment status, tracking links, and lifetime value (LTV).
- Provide quick actions in the panel: Issue Refund, Cancel Unfulfilled Order, and Generate Return Label.

AI RESOLUTION & COPILOT LAYER:
- Integrate an LLM route for AI Auto-Reply and Agent Assistance.
- Implement continuous vector search (pgvector) over uploaded Knowledge Base markdown files. On ticket creation, search relevant context and generate an AI draft response for agent approval or auto-reply if confidence score > 0.88.

OUT OF SCOPE:
- Native telephony/IVR stacks (integrate via generic WebRTC/Twilio embed).
- Complex workforce management (WFM) scheduling.
- Deep historical data migrations from legacy helpdesks.

$ each button prefixes agent-specific run instructions · build your own product, never copy proprietary code, trademarks or designs