Skip to main content

Best Practices

Follow these best practices to build robust, efficient applications with vfrog.ai.

Security​

Protect Your API Keys​

# ✅ Good: Use environment variables
export VFROG_API_KEY="your-key-here"

# ❌ Bad: Hardcode in source
api_key = "vfrog_abc123..."  # Don't do this!

Rotate Keys Regularly​

Rotate API keys every 90 days or when team members leave.

Performance​

Use the Sync Endpoint​

The sync endpoint (/v1/cv/requests/sync) is currently the only available endpoint. It's suitable for both real-time features and batch processing (with appropriate rate limiting).

Optimize Images​

# Resize large images before uploading
from PIL import Image

img = Image.open('large_image.jpg')
img.thumbnail((1024, 1024))
img.save('optimized.jpg', quality=85)

Implement Caching​

Cache results for frequently processed images.

Reliability​

Implement Retries​

from tenacity import retry, stop_after_attempt, wait_exponential

@retry(
    stop=stop_after_attempt(3),
    wait=wait_exponential(multiplier=1, min=4, max=10)
)
def make_api_call():
    return requests.post(url, headers=headers, json=data)

Handle Webhooks Properly​

  1. Return 200 OK immediately
  2. Process asynchronously
  3. Make idempotent
  4. Verify signatures

Monitoring​

Track Usage​

Monitor your API usage in the console and set up alerts at 50%, 80%, and 100% of quota.

Log Everything​

import logging

logging.info(f"Processing image {external_id}")
logging.error(f"Failed to process {external_id}: {error}")

Set Up Alerts​

Configure alerts for:

  • High error rates
  • Quota warnings
  • Failed webhooks

Cost Optimization​

  1. Use appropriate rate limiting for batch processing (10 requests/minute)
  2. Batch process images
  3. Implement caching
  4. Optimize image sizes
  5. Set appropriate quotas per API key