Stories from people already building with AI SDLC

I came to try it for the hype and stayed for the traceability. My requirements used to live in my head and three notes; now I see which test is tied to which requirement. For a side project it's almost overkill in a good way 😅

I don't write code, but I constantly need to turn ideas into proper requirements and user stories. I used to do it by hand for half a day. Here I described a feature, got stories with acceptance criteria and linked them to what already exists. Saved a ton of nerves.

I already have Cursor. The joy is that AI SDLC prepares a proper brief I drop into it — and the code comes out predictable, not "as luck would have it". Plus I recovered the docs I'd given up on back in March straight from the code.

Not everything's perfect — a couple of times I fixed the wording after the AI. But the speed is incomparable: use cases and BPMN from requirements come together in minutes, not an evening. Reviewing and approving — yes, that stays on me, and rightly so.

I'm building solo. The most valuable part is that everything from idea to test cases is linked and versioned. When an investor asked "where's your architecture", I didn't blush. Put together the artifact package over a weekend.

As a newcomer it helps me understand "how software is actually made". I see the whole chain: requirement → use case → code → test. Better than any course, because it's on my own project.

I hand clients a project with documentation I used to hate writing. Now it's recovered from the code and doesn't lie. Natural-language search across the project deserves a separate shout-out — it really finds things by meaning.

I got it for onboarding. A new person joins, opens the project link map and sees what grows from what, without hour-long calls. The traceability matrix turned out more useful than I expected.