Build a tool that uses AI to generate complete software systems, including frontend, backend, and database, while ensuring all components work together reliably.
Over the last few months we've been experimenting with something interesting. Instead of AI generating small pieces of code, we started exploring whether it could build complete software systems — frontend, backend, database, and deployment. At first it sounded straightforward. AI is already pretty good at generating code snippets. But we quickly realized something: generating code isn't the hard part. The real challenge is making all those pieces work together reliably. AI can produce a frontend, an API, and database models. But connecting them into a system that can actually run, deploy, and scale in production is much harder than expected. Most AI tools focus on code generation. But the missing layer seems to be infrastructure orchestration — managing how all those components actually work together. Curious to hear from other builders here: If AI could generate full SaaS products automatically, what part do you think would still be hardest to automate? Architecture? Deployment? Scaling? Would love to hear what others building in this space think.