When we built CreatorCon C3 for Knowledge 2025, we pushed ServiceNow’s attachment system to its absolute limits by processing over 130,000 images from user photos to AI-generated avatars. Along the way, we uncovered hidden quirks, performance bottlenecks, and architectural realities that every developer should know before building anything image-heavy on the platform.

From the two-API “gotcha” for image fields to the sobering math of base64 encoding at scale, this deep dive shares hard-won lessons, practical workarounds, and when you might want to consider a hybrid approach. If you’ve ever wrangled more than a handful of files in ServiceNow, or plan to, you’ll want to read this before your next upload queue grinds to a halt.