Cloud-rendered photo editing for low-end machines

The moment arrives when a freelance photographer or small studio employee opens a layered PSD on a mid-range laptop and watches it stutter through every brush stroke — not because they're doing something unreasonable, but because the software was designed assuming hardware budgets that don't match their reality. Adobe has no structural incentive to fix this: their pricing pushes users toward higher subscription tiers, and optimizing for lower-end hardware would require them to fundamentally re-architect software they've been building for 30 years. That's not a sprint fix.

The specific complaints here aren't about missing features — they're about the compute load. 'Consumes more RAM,' 'requires a machine with high capacity,' 'slow or freeze your computer' — these are resource complaints, not UX complaints. The gap is that all the heavy processing happens locally, on hardware that can't keep up.

What someone can build instead: a thin client that sends editing operations to a remote compute environment, returns a rendered result, and keeps the local experience feeling responsive. The user interacts with a lightweight interface; the GPU-heavy work happens elsewhere. This isn't a new concept for video or 3D, but it hasn't been done in a focused, workflow-native way for still photo editing — especially not for the PSD/RAW editing workflows photographers actually use.

This is a recurring business because photographers don't upgrade laptops every year, but they do edit photos every week. The cost of not solving it is concrete: slower turnaround on client work, inability to take on volume jobs, or spending money on hardware upgrades just to run one piece of software. A monthly subscription that costs less than a RAM upgrade is an easy sell if the latency is tolerable.

What to build

Build a browser-based PSD and RAW editor that offloads all rendering and filter computation to cloud GPUs, streams the canvas back to the client in real time, and supports layer-based editing without requiring any local install beyond a browser tab.

Where to start

Start with batch export and non-destructive adjustment workflows (color grading, exposure, lens correction on RAW files) where near-real-time streaming matters less than it does for brush-based retouching, so you can ship a working product before solving the hardest latency problem.

The hard part

Perceived latency is the killer — even 200ms of round-trip delay on a brush stroke feels broken, so the infrastructure and streaming architecture has to be exceptionally tight before you can demo this to a skeptical photographer without them immediately dismissing it.

How it makes money

Monthly subscription tiered by compute hours used, starting around $15/month for light users and scaling to $50+/month for high-volume editors; no free tier but a 14-day trial with a credit card.