# quirq

> Secure environments for agentic workforces. Any model, any harness, any
> cloud. quirq deploys, manages and meters the agents a team already runs, and
> measures what they delivered rather than what they consumed.

quirq is built on one idea: the token is a good meter for what a machine
consumes and a bad one for what it produces. A quirq is the output unit. A
human owner budgets an outcome, the environment verifies it against captured
state, and the ledger records what the work cost all-in.

This file lists the pages in the site navigation. It is a map, not the corpus:
each page below is the canonical source for its own subject, and the whitepaper
at /whitepaper carries the full construction in prose.

## Pages

- [Products](https://quirq.ai/products): Deploy an agentic workforce
  environment two ways. XO Cloud (Managed) is self-serve and priced per environment;
  bring your own cloud is licensed onto infrastructure you already own and
  stood up by forward-deployed engineers. Covers the harnesses you can deploy,
  the observability layer, the comparison between the two, and pricing.

- [Research](https://quirq.ai/research): The research program behind quirq.
  Experiments, frameworks and field notes across three shelves (Speed Trials,
  From the desk, Proving grounds). Every claim ships with the result that would
  falsify it. Includes the agent-context studies and the replication that did
  not reproduce the pilot's headline.

- [Writing](https://quirq.ai/writing): News, thoughts and guides. Launches and
  integrations, the research findings argued rather than reported, and the unit
  of work explained in reading order.

- [Enterprise](https://quirq.ai/machinespeed): Machine speed. Agentic workflows
  built and run on the stack you already own, with every workflow reporting the
  money it made or saved. The forward-deployed engagement behind a licensed
  deployment.

## Notes for agents

- The whitepaper is readable in full at https://quirq.ai/whitepaper, and the
  typeset PDF at https://quirq.ai/whitepaper/pdf is the version of record.
- The documentation index is https://quirq.ai/docs.
- Correspondence: hello@quirq.ai