Faster media delivery to optimize app performance
Article Summary
LazyPay was serving unoptimized images directly from S3, killing app performance and burning bandwidth. Their migration to Cloudinary cut image sizes by 70%.
The LazyPay engineering team shares their journey from AWS S3 to Cloudinary for media delivery. This case study reveals how they optimized mobile and web performance while reducing maintenance overhead across Android, iOS, and web platforms.
Key Takeaways
- Reduced image payload by 70% using Cloudinary's automatic transformations
- Cut maintenance from 7 image variants per asset to just 1 master
- Achieved 20% faster load times vs S3 through multi-CDN delivery
- Used Firebase Remote Config for controlled rollout and error monitoring
- Created platform-specific named transformations for Android, iOS, and web
LazyPay eliminated 70% of image bandwidth and simplified asset management by replacing direct S3 delivery with Cloudinary's on-the-fly optimization.
About This Article
LazyPay's apps were slow because they weren't handling images well. Users on 4G networks saw sluggish screen rendering, the cache kept missing and causing delays, and the team had to maintain seven different image versions for each asset across Android density buckets (mdpi to xxxhdpi) and iOS scale factors (1x to 3x).
Ashwini Kumar's team put Cloudinary between S3 and their apps to handle image optimization. It uses named transformations to serve the right version for each platform from a single master image and automatically converts formats (WebP, AVIF, PNG) based on what the client can handle.
Cloudinary's multi-CDN delivery and automatic optimization cut image file sizes by 70%. A 100 KB image now gets served at 30 KB. Firebase Performance Monitoring showed image URLs loading 20% faster compared to pulling directly from S3.