Monorepo

So this is like one-to-one relation and this works also perfectly.

But I would be also cautious because when you say monorepo and monobuild, Others can see, for instance, millions of lines of code.

No, not millions.

It's not microservices.

And then essentially these are the next things you tackle because others are already solved problems.

So yeah, the practices, the quality.

And the last thing regarding the monorepo size.

So in a recent project or recent project, also like a legacy migration project, we counted the lines of code.

The Wall Street Journal reported that it took them less than 72 hours to get inside.

They even managed to access the OpenAI GitHub repository, known internally as Monorepo.

Which is a massive deal because Monorepo reportedly houses all the algorithmic secrets for OpenAI.

The researchers did not actually access the internal code themselves, but they proved their access by sending a pull request from a compromised Codex account.

So how exactly did three independent researchers pull this off so quickly against a tech giant? You would think a company building the most advanced models in the world would have ironclad defenses against this exact kind of threat.