Case study

Aether

How Aether introduced its cloud coding agents to 420K+ views

Aether gives a coding agent its own cloud machine and runs it on the subscription you already pay for. The launch film had to explain that to developers in a feed full of agent products, and make the one difference that matters stick.

Views on X
Likes on X
500+
Replies on X
182

Aether × EPTA

Aether gives a coding agent a cloud machine of its own. Hand Claude Code or Codex a task and it boots a machine, runs the work where you can watch every command, and opens the pull request with the proof attached. It runs on the Claude or ChatGPT plan you already pay for, so the only bill is the machine.

Pranav Sharan founded it after building infrastructure at Apple, Uber and HeyGen, and after trying every cloud agent on the market and finding a drawback in each. Aether is the one built so that closing the laptop does not stop the work.

The brief

By November 2026 every developer on X had seen a dozen agent launches, and most of them looked the same: a terminal, a diff, a claim about speed. Aether's difference is where the agent runs and who pays for the model, which is a real advantage and a hard one to dramatise.

The brief was to make one idea, your agent keeps working after you walk away, land in the first ten seconds, without a screenshot of a dashboard doing the talking.

The film

So the film has no founder and no desk. It follows a pizza delivery driver through a shift, laptop open on the delivery bag between drops, handing a task to Aether from the kerb and checking the build from a phone at the next door. The product appears the way it would in life, on the screens in the driver's hands, and lifts into the air as augmented reality when the film needs you to read it.

The line from the post is the whole structure of the film:

Close the laptop. It keeps going.

One minute and forty-four seconds, shot on location, and the agent finishes the job while the driver finishes the route.

Next launch

Planning a launch?

Tell us what is launching and when. Thirty minutes is usually enough to know whether we are the right team for it.