What Presti AI users actually want

8 sources describe specific output quality failures — floating products, color saturation problems, clashing shadows, mismatched reflections — but none indicate users receive diagnostic feedback explaining why these occurred or how to fix them. Users achieve 4x output increases but suffer from low usability rates (2-3 usable out of 4 weekly in manual workflows) when outputs don't meet standards.

4 sources explicitly request AI relighting to generate dozens of variations without reshoots, eliminating additional production cost and time. Users report coordinating sets and props across thousands of products creates speed-to-market delays, and CGI costs $300+ per image with 2-week turnarounds. The request for relighting capability directly addresses the constraint that one product shot should unlock multiple scene variations but currently doesn't.

10 sources confirm users need custom AI style models trained on brand-specific photos to maintain visual consistency across thousands of images. Users explicitly state every generated image must feel like it was styled by their own team, not generic AI. Yet the current evidence suggests custom training happens through Presti team collaboration, not self-service — a bottleneck that limits how fast users can iterate on brand aesthetics.

4 sources explicitly request text generation within images for custom messaging and more engaging content creation. Users deploy AI-generated images across multiple channels (website, lookbooks, social media, sales materials, third-party marketplaces) and need to add promotional messaging, pricing callouts, or campaign headlines without post-processing in external tools.

26 sources show users achieve 4x output increases and generate multiple unique variations with single-click initiation, but no evidence indicates how users organize, review, or select final outputs when generating dozens or hundreds of images per session. The workflow gap: batch generation creates speed, but reviewing outputs becomes a bottleneck when users must manually compare variations across lighting, composition, and angle.

7 sources clarify IP ownership transfers only for downloaded outputs, and users bear full responsibility for input legality and output usage compliance. Yet no evidence indicates users receive automated assistance detecting potential IP violations — a significant legal risk when scaling to thousands of generated images annually. Users report inability to keep visual consistency across marketing channels and third-party marketplaces, suggesting outputs are published widely across surfaces with varying legal requirements.