Introducing NOVEMADE

NOVEMADE is an independent product studio. We design, build, publish and grow our own apps, games and AI products. There were seven of them the day this was written; the products page has the current list, and this site is where they live.

The studio exists because of a specific, unglamorous problem. The first app taught us how to do store assets, subscription plumbing, release checklists, privacy documents, screenshots and analytics. The second app taught us all of it again, from nothing, because none of it had been built to be reused. By the third it was obvious that the repeated work was not a tax on making products — it was most of the work.

What a studio actually buys you

A studio is not a bigger team. It is the decision that the things which are identical across products get built once, and the things that make a product distinctive get the time that frees up. A cozy puzzle game and a sprint-analysis tool have almost nothing in common at the surface. Underneath, they need the same release pipeline, the same creative production, the same measurement, the same legal scaffolding.

So that is what we built underneath them, and it is why the fifth product costs less to launch than the first did. The Technology page describes what that system does, at the level of what it produces rather than how it is deployed.

What we are not

We are not an agency. We do not take briefs, bill hours, or keep a wall of client logos. There is no "Get a quote" button on this site and there is not going to be one, because there is nothing here to quote for. The products are the business.

That is a real constraint rather than a positioning statement. It means no client pays for our mistakes, and it means a product ships when it is genuinely ready instead of when someone approves the invoice.

Where things actually stand

Three products — Didi, SPRIVO Sprint and Luma Jars — are on the App Store, free, since the middle of September 2026. Their Android builds are still in closed testing, so there is no Google Play link on this site for them. ZUPI and Glow Garden are in closed testing on both platforms. Veyra is a working development build with no cloud inference connected. Pivumo has a finished, licence-cleared audio package in twenty-nine languages and an app still in build.

Writing that down plainly is the point. A studio site where every product is triumphantly "live" tells you nothing, because it would look the same either way. The status badges here are accurate, including the ones that are not flattering — and they are re-checked against the stores rather than carried forward, which is how the first version of this paragraph came to say that nothing was public four days after three things were.

Update, 1 October 2026. The paragraph above is left exactly as it was written on 20 September, and it is now wrong in three places. Pivumo went live on 22 September. ZUPI and Veyra both went live on 30 September — Veyra, which that paragraph calls a working development build with no cloud inference connected, is on the App Store at version 3.0.0, and the "no cloud inference" half was already wrong when it was written, which is a separate story recorded in this repository. Two of the three were still wearing a BETA badge on this site the following morning. The store check this article is about is what caught them; nothing in any repository changed to mark the difference, because nothing had to. That is the failure mode, for the third time, and it is why the counts on this site are computed from the catalogue rather than typed into a sentence by anybody — including this one, which names no totals on purpose.

Play something

The most useful thing on this site is not a paragraph. ZUPI’s engine has no third-party runtime dependencies and a clean boundary between its rules and its rendering, which means the demo under Try is not a recreation — it is ZUPI’s actual engine, running level 1 exactly as the app generates it.

Update, 3 October 2026. The paragraph above is false and is left standing because deleting it would hide the most interesting failure on this site. Every word of it was true of the ZUPI that existed when it was written. The ZUPI on the App Store since 30 September is a different game — ZUPI: Pulse Shield, a shield deflecting threats from a core — and the engine that demo ran is the previous one, a balloon under an umbrella. The copy was faithful to the repository the whole time; a hash proved it every build. What no hash could prove is that the repository was still the product. The demo is a recreation now and says so, and the two pieces below about running the real engine carry the same correction.

A studio site should not be a graveyard of things we once made. It should be a window you can reach through.