Every engineering leader is circling the same question right now: how do we move some of our engineering work to the cloud? The keyword is some. Talk to a startup and the question changes to all of it. A few months ago startups only wanted background agents, async tasks running while real work stayed local. Now they want every session in the cloud, synchronous work included, no local dev, laptops reduced to review terminals. Enterprises will get there too, starting with a smaller ask, tag an agent in chat and get a pull request back, but the direction is not in doubt.
What slows the enterprise version down is not the tooling. It is the codebase. Running a coding agent in the cloud means cloning your repositories into the environment where the agent executes, and source code is about the most sensitive IP a company owns. For many security teams, shipping it into a vendor's cloud is a non-starter. That constraint does not kill the shift. It dictates its shape: the agents come to the code, running inside the company's own cloud, rather than the code going to the agents. The one external dependency left is model access, and security teams already have an answer for that in the LLM gateways the big clouds provide, dedicated endpoints like Bedrock and Vertex they have vetted and approved.
The buying pattern here is familiar from a decade of data infrastructure. Most teams prefer SaaS because it is less to operate. The same buyer flips to self-hosted the moment their security team requires it. The engineering team picks the product. The security team picks the deployment model. Any platform that only offers one of the two is invisible to half the market.
This is why cloud dev environments are the unlock for the whole agent story. Parallel sessions, background execution, work that continues while humans sleep, none of it happens on a MacBook. The background layer is still underbuilt, and the companies that build it in a form security teams can approve are the ones that will carry enterprises across.
If you lead engineering, stop debating whether this shift happens and start scoping your version of some. Pick the workflows that move to cloud environments this quarter, and start the security conversation now, because it is the longer pole than the technology.
Key takeaways
- Startups are asking how to move all of their engineering work into cloud agent environments while enterprises are asking how to move some of it.
- The codebase is the most sensitive IP a company owns, so where the agents run is a security decision before it is a tooling decision.
- The security team, not the engineering team, ultimately decides between SaaS and self-hosted, so the winning platform has to offer both.
FAQ
Why do enterprises resist running coding agents in a vendor's cloud?
Because running a coding agent means cloning your repositories into the environment where it executes. Source code is among the most sensitive IP a company holds, and many security teams will not allow it to leave their own cloud, regardless of how good the tooling is.
Will self-hosted always be the enterprise default for agent platforms?
No. Most teams actually prefer SaaS when their security team allows it, because it is less to operate. The pattern is that the same buyer chooses SaaS or self-hosted depending on what security mandates, which is why a platform needs to run credibly in either mode.
Related Essays
Cloud Dev Environments Are the Unlock for Parallel Agent Work
Running agents locally hits a wall fast. Cloud-based dev environments let teams run parallel agent sessions, share work instantly, and actually scale.
The Cowboy Developer Era Is Ending
The solo engineer battling code on a laptop for weeks is a dying model. Units of engineering work are becoming multiplayer, and multiplayer work has to live in the cloud.
Synchronous Cloud Development Is the Next Wave
Three months ago every startup wanted async background agents. Now the leading teams want all their coding sessions in the cloud, synchronous work included.
Drowning in pull requests that need your review? Try Tembo Review, a beautiful AI-assisted PR review tool unlike anything you’ve used.