In a discussion on the topic of #1008, I had pointed out that we should consider moving blocking calls to remote object stores into the background, or access something similar to a sync API that pushes updates from S3 to relevant parseable instances that can then update their state, this would improve latencies concerning API requests.