Skip to main content

Requirements

Before getting started with Brizy AI, ensure you have the following basic requirements and API keys configured.

System Requirements​

Minimum Requirements​

  • Docker: Version 20.10 or higher
  • Internet Connection: Required for API access and image downloads
  • Storage: At least 2GB available disk space
  • Memory: Minimum 2GB RAM for optimal performance
  • Docker Compose: For easier deployment and management
  • Storage: 5GB+ available disk space for templates and assets
  • Memory: 4GB+ RAM for better performance
  • Network: Stable internet connection for API calls

Database (MySQL)​

The AI service (Laravel) stores application data in MySQL. For local development, the AI project’s docker-compose.yml defines a mysql service that matches the default connection settings below.

Purpose: Persistent storage for AI project (jobs metadata, and related application data).

Setup Instructions:

  1. Use Docker Compose (recommended)
    From the AI project root, start the stack so the mysql service is healthy before ai-core starts (depends_on with health checks).
  2. Database and user
    The sample compose file creates database ai_core and allows the root user with an empty password for local dev (adjust for production).
  3. Host from inside containers
    Set DB_HOST=mysql so the app resolves the database service on the Docker network—not localhost from inside the ai-core container.

Environment variables:

VariableDescription
APP_SERVICEService identifier (e.g. ai-core).
DB_CONNECTIONUse mysql.
DB_HOSTHostname of MySQL (mysql when using compose service name).
DB_PORTUsually 3306.
DB_DATABASEDatabase name (e.g. ai_core).
DB_USERNAMEDatabase user (e.g. root in sample compose).
DB_PASSWORDUser password (empty in local sample; set in production).

Message Queue (RabbitMQ)​

AI project uses Laravel’s queue system with RabbitMQ as the backend. Long-running AI work is dispatched to the queue; the ai-core-worker service in docker-compose.yml runs a consumer on the default queue.

Purpose: Reliable asynchronous processing so HTTP requests stay responsive.

Setup Instructions:

  1. Start RabbitMQ
    The AI project includes a rabbitmq service (rabbitmq:3-management) with AMQP on port 5672 and the management UI on 15672.
  2. Credentials
    Align RABBITMQ_USER / RABBITMQ_PASSWORD with RABBITMQ_DEFAULT_USER / RABBITMQ_DEFAULT_PASS in compose (the sample uses admin / admin).
  3. Workers
    Ensure ai-core-worker (or an equivalent php artisan messenger:consume process) is running; it waits for RabbitMQ, then consumes the queue named in RABBITMQ_QUEUE.
  4. Queue driver
    Set QUEUE_CONNECTION=rabbitmq so Laravel uses the RabbitMQ connector, not sync or database.

Environment variables:

VariableDescription
QUEUE_CONNECTIONMust be rabbitmq for this setup.
RABBITMQ_HOSTBroker hostname (Docker Compose service name is usually rabbitmq).
RABBITMQ_PORTAMQP port, typically 5672.
RABBITMQ_USERRabbitMQ username.
RABBITMQ_PASSWORDRabbitMQ password.
RABBITMQ_VHOSTVirtual host (often /).
RABBITMQ_QUEUEQueue name workers consume (e.g. default).

API Keys Setup​

To use Brizy AI features, you'll need these API keys configured:

Required APIs​

1. AWS S3 Bucket​

Purpose: For storing assets, templates, and configurations

Setup Instructions:

  1. Create AWS Account: Sign up at AWS Console
  2. Create S3 Bucket:
    • Go to S3 service in AWS Console
    • Click "Create bucket"
    • Choose a unique bucket name
    • Select your preferred region
  3. Configure CORS: Add CORS policy for web access
  4. Create IAM User:
    • Go to IAM service
    • Create a new user with programmatic access
    • Attach S3 permissions policy
    • Save Access Key ID and Secret Access Key

2. AI Provider (OpenAI or Cerebras)​

Purpose: Large-language-model inference for text generation, structured content, and all related AI features inside project.

AI supports two providers, selected with APP_AI_PROVIDER:

  • openai (default when unset or set to openai): Uses the OpenAI API. Pair with APP_OPENAI_API_KEY and optionally APP_OPENAI_ORGANIZATION. Fine-tuned and model-specific IDs are configured via the APP_GPT_MODEL_* variables (see the environment example below).
  • cerebras: Uses the Cerebras inference API. Set APP_AI_PROVIDER=cerebras and APP_CEREBRAS_API_KEY.

Choosing Cerebras

ProsTypically lower latency / faster generation for supported models—useful when throughput or responsiveness matters.
ConsPricing and quotas differ from OpenAI; cost at scale should be validated against your workload and budget.

You only need credentials for the provider you enable; the other provider’s keys can remain empty.

OpenAI — setup:

  1. Create OpenAI Account: Sign up at OpenAI Platform
  2. Get API Key:
    • Go to API Keys
    • Click "Create new secret key"
    • Copy and save the key securely
  3. Organization (optional): Set APP_OPENAI_ORGANIZATION if the key belongs to an organization.
  4. Add Payment Method: Required for API usage beyond free trial limits
  5. Check Usage Limits: Monitor your usage in the dashboard

Cerebras — setup:

  1. Obtain an API key from Cerebras (per current account and product docs).
  2. Set APP_CEREBRAS_API_KEY and APP_AI_PROVIDER=cerebras.

Note: OpenAI may offer limited free credits for new accounts; always check the latest pricing for both providers.

3. Google Places API Key​

Purpose: For location-based data and search functionality

Setup Instructions:

  1. Create Google Cloud Account: Sign up at Google Cloud Console
  2. Create Project: Create a new project or select existing
  3. Enable Places API:
    • Go to APIs & Services > Library
    • Search for "Places API"
    • Click "Enable"
  4. Create Credentials:
    • Go to APIs & Services > Credentials
    • Click "Create Credentials" > "API Key"
    • Copy the generated key
  5. Restrict Key (Recommended):
    • Click on the created key
    • Add restrictions for security

Note: Free tier includes 100 requests per day

4. Facebook API Key​

Purpose: For social media integration and user data

Setup Instructions:

  1. Create Facebook Developer Account: Go to Facebook Developers
  2. Create App:
    • Click "Create App"
    • Choose "Business" type
    • Fill in app details
  3. Get App Credentials:
    • Note down App ID and App Secret
    • Configure OAuth settings if needed
  4. Set Permissions: Configure required permissions for your use case
  5. API version: Set APP_FACEBOOK_API_VERSION (e.g. v23.0) to match your app.
  6. Instagram (optional): If you use Instagram-related features, configure APP_INSTAGRAM_ID and APP_INSTAGRAM_TOKEN.

Note: Free tier includes 100 requests per day

5. Unsplash API Key​

Purpose: For high-quality stock images

Setup Instructions:

  1. Create Unsplash Account: Sign up at Unsplash
  2. Register as Developer:
  3. Create Application:
    • Provide app name and description
    • Set traffic source URL
    • Submit for approval
  4. Get API Key: Copy the Access Key from your application dashboard

Note: Free tier includes 50 requests per hour

Environment Configuration​

Required Environment Variables​

After obtaining your API keys, you'll need to configure these environment variables:

# AWS S3 Configuration
AMAZON_S3_REGION=your-bucket-region
AMAZON_S3_BUCKET=your-bucket-name
AMAZON_S3_KEY=your-access-key-id
AMAZON_S3_SECRET=your-secret-access-key

# Fine-tuned / model IDs when using OpenAI only (omit if cerebras)
APP_GPT_MODEL_HEADING=your-openai-model-id
APP_GPT_MODEL_HEADING_LOCATION=your-openai-model-id
APP_GPT_MODEL_SUBHEAD=your-openai-model-id
APP_GPT_MODEL_SUBHEAD_HEADING=your-openai-model-id
APP_GPT_MODEL_SUBHEAD_LOCATION=your-openai-model-id
APP_GPT_MODEL_SUBHEAD_HEADING_LOCATION=your-openai-model-id
APP_GPT_MODEL_TESTIMONIAL=your-openai-model-id
APP_GPT_MODEL_SERVICE_NAME=your-openai-model-id
APP_GPT_MODEL_SERVICE_DESCRIPTION=your-openai-model-id
APP_GPT_MODEL_TRANSLATION=your-openai-model-id
APP_GPT_MODEL_FROM_REVIEWS_TO_BUSINESS_DESC=your-openai-model-id
APP_GPT_MODEL_BUSINESS_NAMES=your-openai-model-id
APP_GPT_MODEL_BUSINESS_TYPE=your-openai-model-id

## OPENAI
APP_OPENAI_API_KEY=your-openai-api-key
APP_OPENAI_ORGANIZATION=your-openai-org-id

## CEREBRAS
APP_CEREBRAS_API_KEY=your-cerebras-api-key
APP_AI_PROVIDER=cerebras

## GOOGLE PLACES
APP_GOOGLE_PLACES_API_KEY=your-google-places-key

## FACEBOOK
APP_FACEBOOK_ID=your-facebook-app-id
APP_FACEBOOK_CONFIG_ID=your-facebook-config-id
APP_FACEBOOK_SECRET=your-facebook-app-secret
APP_FACEBOOK_API_VERSION=v23.0
APP_INSTAGRAM_ID=your-instagram-id
APP_INSTAGRAM_TOKEN=your-instagram-token

## UNSPLASH
APP_UNSPLASH_API_KEY=your-unsplash-access-key
APP_UNSPLASH_TRAFFIC_SOURCE_URL=your-website-url

# URL to API for blocks
APP_BLOCKS_API_URL=

# URL to validate API keys (your gateway endpoint)
APP_VALIDATE_API_KEY_URL=

## DB
APP_SERVICE=ai-core
DB_CONNECTION=mysql
DB_HOST=mysql
DB_PORT=3306
DB_DATABASE=ai_core
DB_USERNAME=root
DB_PASSWORD=

## QUEUE
QUEUE_CONNECTION=rabbitmq

RABBITMQ_HOST=rabbitmq
RABBITMQ_PORT=5672
RABBITMQ_USER=admin
RABBITMQ_PASSWORD=admin
RABBITMQ_VHOST=/
RABBITMQ_QUEUE=default

Use APP_AI_PROVIDER=openai when using OpenAI; use cerebras with APP_CEREBRAS_API_KEY as shown above. Only one provider is active at a time.

Quick Start Checklist​

Before proceeding to the Usage Guide, ensure you have:

  • Docker installed and running
  • MySQL available and DB_* variables set
  • RabbitMQ running (e.g. rabbitmq service healthy) and QUEUE_CONNECTION=rabbitmq with matching RABBITMQ_* values
  • queue worker running if you rely on async jobs
  • AWS S3 bucket created and configured
  • AI provider configured: OpenAI (APP_OPENAI_API_KEY, optional org, APP_GPT_MODEL_*) or Cerebras (APP_CEREBRAS_API_KEY, APP_AI_PROVIDER=cerebras)
  • Google Places API key created
  • Facebook API credentials ready (and Instagram vars if needed)
  • Unsplash API key registered
  • All environment variables documented
  • Network access to all required services

Next Steps​

Once you have your API keys ready, proceed to the Usage Guide to learn how to integrate Brizy AI into your project.

Troubleshooting​

Common Issues​

API Key Errors:

  • Verify keys are correctly copied (no extra spaces)
  • Check if keys have proper permissions
  • Ensure services are enabled in respective dashboards

S3 Access Issues:

  • Verify bucket name and region match
  • Check IAM user permissions
  • Ensure CORS is properly configured

Rate Limiting:

  • Monitor usage in respective dashboards
  • Implement proper error handling for rate limits
  • Consider upgrading plans if needed

Database connection errors:

  • Confirm DB_HOST matches the Docker service name from the app container’s network
  • Ensure MySQL finished startup (health check passed) before ai-core requests the DB

Queue / RabbitMQ issues:

  • Verify QUEUE_CONNECTION=rabbitmq and broker host/port from inside the worker container
  • Confirm the worker process is running and consuming RABBITMQ_QUEUE
  • Use the management UI on port 15672 (when enabled) to inspect queues and consumers

For additional support, see the API Reference or contact our support team.