Can I vibecode Klaviyo?

Build a detailed, self-hosted clone inspired by Klaviyo focusing on its core commerce marketing and automation capabilities. Use exactly this stack: Next.js 15 (App Router), TypeScript, Tailwind CSS, PostgreSQL, and Drizzle ORM.

**DATA MODEL & SCHEMA SPECIFICS**
Implement the following core entities in PostgreSQL with Drizzle ORM:
- `profiles`: `id`, `email`, `first_name`, `last_name`, `properties` (JSONB).
- `events`: `id`, `profile_id`, `event_name` (e.g., 'Placed Order', 'Viewed Product'), `event_data` (JSONB), `timestamp`.
- `segments`: `id`, `name`, `rule_definition` (JSONB).
- `campaigns`: `id`, `subject`, `html_body`, `status`, `target_segment_id`.
- `flows`: `id`, `name`, `trigger_event`, `flow_definition` (JSONB - nodes and edges for automation steps).
Ensure all relations have cascading deletes where appropriate. Use foreign key constraints.

**SCREENS & UX PATTERNS**
1. **Audience & Segmentation (`/audience`)**: A data table of profiles. Includes a query builder interface (using compound AND/OR logic UI) to create dynamic segments based on historical events (e.g., "Placed Order at least once in 30 days").
2. **Flow Builder (`/flows/[id]`)**: A node-based visual canvas (using a library like React Flow) to map out automation sequences (Trigger -> Delay -> Condition -> Email).
3. **Campaign Editor (`/campaigns/[id]`)**: A drag-and-drop visual email builder optimized for e-commerce, allowing dynamic insertion of product grids based on profile properties.
4. **Analytics Dashboard (`/dashboard`)**: Time-series charts showing revenue attributed to flows vs campaigns, powered by SQL aggregations over the `events` table.

**STEP-BY-STEP USER JOURNEYS**
- **Journey 1 (Data Ingestion)**: User sets up a webhook endpoint. External commerce events (like 'Checkout Started') hit the endpoint and are logged into the `events` table, automatically upserting the corresponding `profile`.
- **Journey 2 (Segmentation)**: User uses the visual query builder to create a "VIP Customers" segment, defining rules that filter profiles where total spent > $500.
- **Journey 3 (Automation)**: User builds an Abandoned Cart flow on the canvas, setting a trigger for 'Checkout Started', a 4-hour delay node, and an email node containing dynamic cart data.
- **Journey 4 (Campaign Send)**: User designs a promotional email, selects the "VIP Customers" segment, and schedules the blast.

**TECHNICAL SPECIFICATIONS & PROTOCOLS**
- Use Server Actions for all form submissions and data mutations.
- E-commerce Data Mocking: Provide a built-in utility script to seed realistic e-commerce profiles and event streams.
- Automation Engine: Implement a robust background job queue (e.g., using BullMQ or simple Postgres-backed queues) to process flow delays and evaluate segment rules asynchronously.
- External Integrations: Standard SMTP protocol integration using Nodemailer for sending outgoing emails.
- Use JSONB heavily for flexible event properties and rule definitions.

Start from an empty folder, create the complete project, and ensure the default mode is single-user and private. Do not add analytics or telemetry. Put all secrets in `.env`. Validate untrusted input with Zod. Write unit tests for the rule evaluation engine and one end-to-end smoke test using Playwright.