The problem
Five idle machines and one that renders all night.
The render sits on one workstation for six hours while the other machines in the building do nothing. Everybody knows this. The reason it stays that way is that the answer has always been a render manager: a server to set up, licences to buy, a service to maintain, and an administrator who is usually you.
For a small structure that overhead costs more than the six hours it saves, so nobody installs it, and the workstation keeps rendering all night.
The answer
The shared folder is the server.
Presto has no central component. Coordination happens entirely inside a folder every machine can already see, through atomic file renames, which is the one operation a network share can be trusted to perform without ambiguity.
Submit a comp and it is split into parts. Every machine running Presto claims the work it can take. One of them combines the finished segments back into the final file. Nothing was installed, nothing is listening on a port, and there is no server to be down.
How it works
The fast machines take more work.
Machines are ranked on cores, RAM and disk speed, then that ranking corrects itself over time against the render speed each machine actually delivers. The workstation that looks powerful on paper but sits behind a slow share drifts down the order on its own.
Each machine can act as renderer, combiner, notifier, or any combination, toggled independently. Mac and Windows machines finish each other's jobs.
What is in it
The rest of the toolbox.
Parallel comp splitting
Any After Effects comp is split into parts and rendered simultaneously across the network.
No server
Coordination lives in the shared folder, through atomic renames. Nothing to install, nothing to keep running.
Self-correcting ranking
Cores, RAM and disk speed to start with, then adjusted against the render speed each machine really achieves.
Local render cache
Render to the local disk and upload the finished segment, which spares the NAS during the render itself.
Live network monitor
A standalone HTML dashboard, readable in any browser on any screen in the building.
Independent roles
Render, Combine and Notify toggled per machine, so a workstation in use can still contribute.
Cross-platform combine
Mac and Windows machines pick up and finish each other's jobs.
Watchdog and retries
An automatic retry budget, a stalled-render watchdog, and kill or restart on any individual part.
Phone notifications
A push on delivery, and a push on failure.
Persistent logs
A per-machine log kept on disk, for working out afterwards what went wrong and where.
Snapshot submission
The render works from a frozen copy of the project, so you can keep editing the original.
Availability
Not distributed publicly.
Built for an internal broadcast workflow, not distributed publicly at this time.