Every product here is in beta. That is a real stage, not a disclaimer — so this page separates what you can verify yourself from what still needs a pilot to establish.
We are on the lower rungs. That is what beta looks like, and pretending otherwise would be the fastest way to lose a technical evaluation.
| Level | Evidence | Where we are |
|---|---|---|
| 1 | Working demonstration | Yes — playground, taskboard example |
| 2 | Public repository | Yes — NC, NC AI, NC UI public; others on request |
| 3 | Reproducible test | Yes — 114-test suite, differential engine test |
| 4 | Published benchmark | Partial — only figures we measured ourselves |
| 5 | Pilot | Not yet |
| 6 | Design partner | Not yet — open now |
| 7 | Customer deployment | Not yet |
| 8 | Published case study | Not yet |
One command. No account. Confirms the single-binary and no-dependency claims in about a minute.
A working backend and interface in about 40 lines. Confirms the development-complexity claim.
Generation still works. Confirms the local-inference claim directly.
We publish a figure only when we measured it ourselves under a stated method, and we say what that method was. Where a number has not been measured that way, the page says so instead of estimating.
The missing rungs need a real problem in a real environment. That is what the design partner programme is for — the engagement produces the evidence, rather than requiring it up front.
Everything at levels one to three is available right now, with no conversation required.