Production Web Application

I need to create a PostgreSQL database structure for my RAG food application. Please help me create:

**Database Requirements:**
- Table name: food_items
- Columns: id (SERIAL PRIMARY KEY), text (TEXT NOT NULL), region (VARCHAR(100)), type (VARCHAR(50))
- Additional timestamp columns: created_at, updated_at (both TIMESTAMP WITH TIME ZONE)
- Default values for timestamps using NOW()

**Tasks:**
1. Create the SQL CREATE TABLE statement
2. Create a Python script using psycopg3 to connect and create the table
3. Create a data migration script to load existing food.json data into the database
4. Add proper error handling and connection management
5. Include a verification script to check the data was loaded correctly

**Database Connection:**
- Use the DATABASE_URL environment variable from Vercel/Neon
- Include SSL mode configuration for Neon
- Handle connection pooling for production use

**Data Migration:**
- Load data from existing foods.json file  
- Preserve all existing data structure (id, text, region, type)
- Add created_at timestamps for existing records
- Handle duplicate entries gracefully
- Provide progress feedback during migration

Please create production-ready Python scripts that I can run locally to set up and populate the database.

Create a comprehensive database migration system for my RAG food application:

**Current State:**
- I have a foods.json file with food data
- New Neon PostgreSQL database with food_items table created
- Need to migrate from JSON file to database

**Requirements:**
1. **Migration Script**: Load foods.json data into food_items table
2. **Verification Script**: Confirm all data migrated correctly  
3. **Backup Script**: Create JSON backup from database
4. **Update Vector Embeddings**: Re-embed from database instead of JSON

**Python Dependencies:**
- Use psycopg3 for PostgreSQL connection
- Support for environment variables via python-dotenv
- Proper error handling and logging
- Progress bars for long operations

**Database Schema:**
```sql
CREATE TABLE IF NOT EXISTS food_items (
  id SERIAL PRIMARY KEY,
  text TEXT NOT NULL,
  region VARCHAR(100) NOT NULL,
  type VARCHAR(50) NOT NULL,
  created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(),
  updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW()
);
```

**Expected foods.json Structure:**
```json
[
  {
    "id": "1",
    "text": "A banana is a yellow fruit that is soft and sweet.",
    "region": "Tropical",
    "type": "Fruit"
  }
]
```

Create scripts that handle edge cases, provide clear feedback, and ensure data integrity throughout the migration process.

I need to create an admin dashboard for my Next.js RAG application with the following features:

**Current Architecture:**
- Next.js 15.5+ with React 19 and Server Actions
- Neon PostgreSQL with food_items table
- Upstash Vector Database for embeddings
- Groq for LLM inference

**Required Admin Features:**

**1. Food Items Management:**
- List all food_items with pagination and search
- Add new food items with form validation
- Edit existing items with inline editing
- Delete items with confirmation dialogs
- Bulk operations (delete multiple, export to JSON)

**2. Vector Database Management:**
- Display current vector count from Upstash
- "Flush All Vectors" button with confirmation
- "Re-embed All Items" - read from PostgreSQL and create vectors
- "Embed New Items Only" - embed items created after last embed date
- Progress tracking for bulk embedding operations

**3. RAG Analytics Dashboard:**
- Query statistics (most asked questions, response times)
- Database health metrics (connection status, query performance)  
- Vector database metrics (total vectors, last update time)
- Usage analytics (queries per day, popular food types)

**Technical Requirements:**
- Use Server Actions for all database operations
- Implement proper error handling and user feedback
- Add loading states for long operations
- Responsive design with Shadcn UI components
- Role-based access (admin authentication)

**Database Integration:**
- Connect to Neon PostgreSQL using DATABASE_URL
- Create additional tables for analytics if needed
- Maintain referential integrity between food_items and vectors

Create a comprehensive admin dashboard that provides full control over the RAG system's data and vector embeddings.