From prototype to product

The difference between a prototype and a product

A prototype shows on your own machine that the idea works. A product is software that someone else can install, whose usage is written down, where it is known what changed in each version, and which can be rolled back when it breaks. The work in between is dull but decisive: packaging, the install path, version numbers, tests, documentation and a way to report bugs. This service takes on that work. The starting point is often a script, a folder or a tool that works on one person's computer. We first try to install it from scratch in a clean environment; every step that fails is a step where a user would also get stuck. We list every gap we find and choose together with you which ones must be closed before the first release.

What we do

• Packaging and versioning: We make the software installable and set up a version numbering routine.
• Setup guide: We write a step-by-step document a user can follow to install and run it from scratch; we do not put a command in the document without trying that it works.
• Release notes: For each version, what is new, what was fixed and what to watch for, kept in date order.
• Test gates: Automated tests and build checks before release; if a gate is red, nothing is released.
• Release pipeline: Depending on the type of software, npm, crates.io, an install script or your own server.

Writing the boundary clearly

What a product does not do should also be in the documentation. When it is written which operating systems it was tried on, which uses it was not designed for and what is known to be missing, users do not install it with the wrong expectations, and you are protected from unexpected support requests. We follow the same method for our own products; our product pages carry only verified information and clear limits. We consider a document written when someone who has never seen the product can read it and install it. Where possible we run this test with a person who did not write the product, because the author does not notice missing steps.

Who owns the source and the release

Release accounts, repositories and keys are opened in your name, so the product does not depend on a single team. The software licence and code transfer terms are settled in writing before work starts; whether it will be published as open source is also decided at that stage.

What determines time and price

Price and time depend on the state of the prototype, the number of operating systems and environments to support, whether tests already exist, the volume of documentation to write, the type of release pipeline and the security-related features of the software. We first review the code and produce a list of findings; the scope is written from that list. The quote depends on the written scope and does not commit you.

If the prototype itself has not been written yet, you can start with custom software development. If you are going to connect an AI model to your application, see the AI integration page.