Fleet is a framework. Pagerox is a workforce.
LangChain Fleet gives engineering teams serious building blocks for running fleets of agents: orchestration, tracing, evals. If you have engineers to spend, you can build something great with it. Pagerox starts where the build ends: the employee already exists, learns your company, and answers to a manager, not a maintainer.
| LangChain Fleet | Pagerox | |
|---|---|---|
| Who operates it | Your engineers: build, deploy, monitor, patch. | Your managers: hire, correct, approve. No code, no ops rotation. |
| Knowledge | You wire retrieval and keep it current per agent. | One org brain with provenance; every employee draws from it role-scoped. |
| Learning | You design the feedback loop, then maintain it. | Anyone corrects in Slack; corrections become attributed, revertible memories. |
| Where it lives | Behind an API you expose to your surfaces. | In Slack, with a name, a face, and an AI badge, where the questions already are. |
| Guardrails | Whatever you implement. | A permission ladder per operation, approvals with digest-verified args, a flight-recorder trace per turn. |
| Time to a working teammate | A sprint, then iteration. | Minutes: hire from a template, connect tools, teach the first correction. |
| Cost shape | Engineering time plus infrastructure plus tokens. | A salary: $99 per employee per month, usage included. |
When to use them instead of us. If agents are your product, or you need full control of every token and tool call, build on Fleet: that is what it is for. Hire from Pagerox when agents are not your product and you just need the work done, accountably, this week.