Pixes: In-Browser Client Privacy Image Suite
Smart, high-performance image manipulation tools executing 100% locally in user browser memory. Convert, compress, resize, crop, and batch process without sending a single byte to remote servers.
The Problem
Almost all mainstream online image conversion and compression tools (e.g. TinyPNG, ILoveIMG) require users to upload confidential documents, sensitive photos, and corporate media to cloud backend servers for processing.
This model causes significant latency (waiting for multi-megabyte uploads and downloads), introduces data retention and GDPR compliance risks, and incurs heavy recurring server GPU/CPU compute costs for the provider.
Architecture & Execution Pipeline
Pixes replaces the traditional backend processing server entirely. Heavy encoding and decoding algorithms (WebP, AVIF, MozJPEG, PNGquant, SVG sanitization) are compiled into WebAssembly (Wasm) binaries and executed asynchronously inside multithreaded Web Workers.
Because all processing happens on the user's multi-core CPU, bulk batches of 50+ images can be converted concurrently in parallel without network bottlenecks.
Performance & Engineering Highlights
• Instantaneous Execution: Eliminating network upload round-trips reduces processing time from ~8 seconds on cloud tools to < 300ms locally.
• Static Edge Hosting: The entire application is served globally via Cloudflare Pages edge CDN, delivering sub-20ms initial page load times.
• Infinite Scalability: Scaling from 10 to 1,000,000 users incurs zero backend compute scaling costs.