← Back to essays
·2 min read·By Ry Walker

The Orchestration Bet

The Orchestration Bet

One of the more interesting architectural decisions in this space is whether to build your own coding agent or to build the layer that sits above all of them. Most companies are betting on building the best agent. A smaller number are betting that the best agent changes every week — and that the durable value is in the orchestration.

If you build your own agent, you are betting that your model integrations, your prompt engineering, and your tool-calling architecture will remain competitive against every other team doing the same thing, including teams backed by billions of dollars. That is a hard bet to sustain. Every quarter, someone ships something better, and the gap between you and the frontier widens by another model release.

If you build the orchestration layer — the system that lets an organization use Claude Code today, Codex tomorrow, and whatever leapfrogs both of them next month — you are making a different bet. You are betting that flexibility and interoperability matter more than any single agent's capabilities at any single point in time. You are betting on the customer's revealed preference, which is to use the best tool for each job and to switch when something better ships.

This is the same architectural insight that has played out in every previous infrastructure wave. The companies that won in cloud were not the ones that built the best virtual machine. They were the ones that built the management and orchestration layers that let enterprises use whatever compute they needed without being locked in. Kubernetes outlasted a dozen better schedulers. Terraform outlasted every cloud-specific tool.

The coding agent that is best today will not be best in 90 days. The platform that lets you swap between them without rearchitecting your workflows — that has a longer shelf life. I've argued elsewhere that the zero-stickiness problem makes any tool-layer position structurally fragile. Orchestration is the answer to that fragility, because it monetizes the act of switching itself.

Key takeaways

  • Building the best single agent is a bet you make against every team in the world doing the same thing, including teams with billions of dollars.
  • Building the orchestration layer is a bet that flexibility and interoperability outlast any single agent's capabilities.
  • This is the same architectural insight that decided the cloud wars — the management layer outlasts the underlying compute.

FAQ

Doesn't owning the agent give you tighter integration?

Yes, but tighter integration with a thing that gets replaced every 90 days is a depreciating asset. The integrations you build today are against a moving target. Orchestration is integration with the abstraction, not the implementation.

What do you actually orchestrate?

Routing tasks across models and harnesses, managing context handoff, enforcing policy and budget, observing execution, and translating between APIs. The orchestration layer is whatever a customer would otherwise build to make multiple agents usable across their org.