What is an AI software factory?

An AI software factory is a production line for software changes. It takes specifications the way product writes them, produces the code and the tests, verifies its own work with an independent check, and stops for human approval before anything merges.

An AI software factory is a production line built on your codebase. It takes specifications written by product, produces the change in the right layer of your system, generates functional and visual tests derived from the acceptance criteria, verifies its own work with an independent check, and stops for human approval. The output lands in your repository and runs in your CI. That is the definition yaab operates by, and every part of it is doing work: what goes in, what comes out, who checks it, and who approves it.

Why "factory" and not "assistant"

AI coding assistants made individual developers faster. The cost per change stayed flat anyway, because every change still needs someone to interpret the requirement, place it in the codebase, write the tests, and review all of it. A factory moves that whole sequence into a pipeline: the pipeline produces the code, the tests, and the first round of verification, and engineering time goes to deciding what to build and approving what lands. The unit of production stops being developer hours and becomes verified changes.

What goes in

The input is a specification the way product already writes it: what should happen, the acceptance criteria, and the scenarios. It describes behavior. The factory decides the implementation, reading your architecture before touching it so the code lands where your team would have put it, following your conventions.

What comes out

For each spec, the factory produces the change in the right layer of your system, functional tests where each acceptance criterion maps to a test that exercises real behavior including the failure paths, visual tests where the change touches the interface, a traceability record of which requirement produced which file and which test covers it, and the run log of what was produced, what was verified, and what failed and was corrected along the way.

Who checks it, and who approves

Two controls make the output trustworthy. First, an independent check: every test the factory produces is reviewed by a verifier that did not write it, looking for assertions loosened, steps skipped, and validations emptied. Second, a human gate: nothing merges without an engineer's approval, inside your existing pull request process.

Frequently asked questions

What is an AI software factory in one sentence?

An AI software factory is a production line on your codebase that turns product specifications into code, tests, and evidence, verified by an independent check and approved by an engineer before anything merges.

Is an AI software factory the same as using AI coding assistants?

An assistant accelerates a developer inside each task. A factory runs the whole sequence, from specification to verified change with tests and a traceability record, and reserves human time for deciding what to build and approving what lands.

Does anything merge without human review?

Nothing merges without an engineer's approval. The human gate runs inside your existing pull request process.

Who owns the code an AI software factory produces?

With yaab, the client owns everything: code, IP, and repository. The factory works in your repository and your CI, so the output is yours from the first commit.