How should a large scale application design its file upload architecture to handle high volumes of uploads reliably, including virus scanning and offloading storage to a dedicated service?

A robust large scale file upload architecture typically offloads actual file storage to a dedicated object storage service rather than the application server's own local disk, which does not scale well across multiple servers, integrates automated virus and malware scanning on every uploaded file before it becomes accessible to other users, processes uploads asynchronously through a background job queue for tasks like generating image thumbnails or extracting document metadata, and enforces strict validation and rate limiting to prevent abuse of the upload feature.

// Simplified flow
// 1. Validate and temporarily store upload
// 2. Queue background job for virus scan and processing
// 3. Move to permanent object storage only after passing all checks