The reason India doesn’t have enough awesome developers

In my experience, this article seems to be accurate. To clarify, I've been working in India as a lead developer for around nine months now. I will be here another three months at the least before returning to the US.

The hiring practices around here for developers is standard for most jobs in the US (from what I'm told). Students from top schools get recruited right from the university if they have good grades.

However, as you may guess, grades aren't always an indicator of quality. Of the seven developers that had been hired since the time I was hired, only around three have any code quality. I've seen front-end developers struggle writing simple Javascript. A backend developer with supposedly three years of experience as a lead can barely write coherent Java, always scouring stackoverflow to find snippets. That same developer also had tried to steal code from me and claim it as his own. It was a git-deployment tool, and given that he never learned to use Git, I was immediately suspicious. Looking through the repository he had created, most of it was copy-and-pasted code with the author attribution overwritten with his name. Thank god for the logs at that moment!

To clarify, I have also seen some excellent developers here. They work extremely fast, and so their work-ethic is enviable; exceed expectations and then work at east on the next features given. They rapidly implement changes in UI and features at a pace I no doubt could not do.

Overall, I agree with this article's point. Maybe myself and the author are wrong, and I just have a bias based on my experience.

When you hire en-masse, it's virtually impossible to test for quality. A lot of these large firms (and major US consultancies follow this too) wind up waving 1000s of people in, and then relentlessly culling those who don't perform. The industry winds up with a lot of turnover too.

The strange thing about this system is that while individual people can underperform, if there's a lot of weeding out (let's say 1 of 5 makes manager, 1 of 25 makes partner) you can wind up with some exceptional people at the end. That's part of why the partners at many consulting firms can achieve so much relatively early in their careers.

Of course this type of arrangement requires a lot of process, and is better for well defined tasks without algorithmic complexity (say building a custom telecom billing system) but it does work in some places.