# How a Google engineer became staff at the age of 28 only

![Image](https://upload.cafenono.com/image/slashpagePost/20250106/203744_omJNl152hQmOu141lW?q=80&s=1280x180&t=outside&f=webp)

I remember the times when I was completely sold on the idea of the Repository pattern. I wrote about it, talked about it in meetups, and used it extensively in projects, especially when I was diving deep into Domain-Driven Design (DDD) and layered architectures. The core principle was always the same: keep the domain layer free of data concerns and push data interactions out to dedicated repositories. It sounded great in theory and worked well for a while in practice.

But along the way, I found myself spending more and more time mapping data between different layers — database entities to domain objects, domain objects to DTOs, and so on — without seeing much real value, especially for typical line-of-business applications. My “Aha!” moment came when I realized how much simpler things could be if I just cut out the middleman, so to speak.

I remember the times when I was completely sold on the idea of the Repository pattern. I wrote about it, talked about it in meetups, and used it extensively in projects, especially when I was diving deep into Domain-Driven Design (DDD) and layered architectures. The core principle was always the same: keep the domain layer free of data concerns and push data interactions out to dedicated repositories. It sounded great in theory and worked well for a while in practice.

But along the way, I found myself spending more and more time mapping data between different layers — database entities to domain objects, domain objects to DTOs, and so on — without seeing much real value, especially for typical line-of-business applications. My “Aha!” moment came when I realized how much simpler things could be if I just cut out the middleman, so to speak.

For the site tree, see the [root Markdown](https://slashpage.com/slashslimes-site.md).