Hacker News new | comments | show | ask | jobs | submit login
[flagged] 95% engineers in India unfit for software development jobs: study (thehindubusinessline.com)
66 points by Etheryte 239 days ago | hide | past | web | favorite | 37 comments

While the 95% claim might be somewhat accurate, I would not trust any studies by Aspiring Minds, which conducts the AMCAT employability test. My college's training and placement department recently forced everyone to answer it ("for practice"), and the quality of the questions was surprisingly poor, from English questions that were grammatically incorrect themselves, to computer science questions with all options incorrect and the worst: computer programming questions about obscure C# APIs (along the lines of "what are the parameters and the return type"), a language which pretty much no one answering the exam has ever learnt.

It's also almost exclusively used by the kind of third-grade mass-recruiting consulting companies that have given Indian developers a bad name. Which is not a surprise, since colleges (including mine) have picked up on this, and now specifically train students for such exams. Even for technical roles (supposedly, although I think spamming StackOverflow hardly qualifies), the online test rarely contains a computer science / computer programming component, and new recruits are commonly expected to undergo year-long training at their own expense even after being hired.

My dislike for Aspiring Minds is hard to put into words. It's strong enough that I would refuse to work for a company that uses them, out of fear of the unchecked incompetence of my future coworkers.

Hard to trust a private study conducted by a company whose interest is in convincing businesses that they can't effectively hire without their tests and certifications.

Even if the figure is true, this is not the right starting point for discussion. Also, we can do better with direct links to studies, rather than bad journalism digests.

It sounds like it may be worse than that. According to someone who took the test[1] they then also offer students the chance to pay to retake for better scores.

[1] https://www.reddit.com/r/programming/comments/66kb3c/95_engi...

If you hire a sweatshop you should expect a sweatshop quality. In a lot of cases hiring these sweatshops (TCS,Wipro,Tech Mahindra etc) is just to save money that can be then paid as bonuses to useless managers.

Having worked for TCS, I 100% agree (US base of operations in Milford Ohio). I gtfo of there in 1.5 yrs, but holy hell was it a shit show. The thing I learned is that many businesses will not respect the iron triangle. They will hire the low bidder, expect it to be done quickly and beat the quality out of them. TCS was all over that kind of work.

It's a sad thing because it's damaging serious businesses like telcos,banks,insurance companies. But I guess the wheel is still in motion and we can keep many people employed.

> It's a sad thing because it's damaging serious businesses like telcos,banks,insurance companies.

Doesn't matter if it's damaging to the business in the long run, it makes their quarterly results look great, and that's all Wall Street cares about. Headcount is down, profit is up.

Who cares if their business logic will be a hot mess in 2 years because everyone competent got fired? Everyone high up will have already collected their performance bonus.

Far too many companies are afraid to take long term decisions which are better for the company but worse in the short term. This was cited as one of the primary reasons for Dell to go private again:

"My partners at Silver Lake Management and I successfully took Dell private a year ago in the largest corporate privatization in history. Iā€™d say we got it right. Privatization has unleashed the passion of our team members who have the freedom to focus first on innovating for customers in a way that was not always possible when striving to meet the quarterly demands of Wall Street." [0]

[0] https://www.wsj.com/articles/michael-dell-going-private-is-p...

I've worked with low-level Indian programmers, and they did write some crazy code. There are also really good ones.

I refuse to believe this article, though. It seems very much a clickbait title/article to get visitors to their website. Any study can prove pretty much anything.

I'd be curious to see the success rate for US-based engineers as well. I suspect it would be higher, but not by much at all. In my experience, outside of the prestige companies and communities, most of the people in our field are dangerously unskilled.

This is not a US vs India comparison. It is a "controlled" data set study done on Indian engineers trained in India. I am sure a separate study can surface interesting trends about US workforce as well.

Exactly! I suspect most American programmers are woefully unprepared for even the most basic of programming tasks. I wouldn't be surprised if that figure - if it can even be credibly measured - is north of 90%.


This is a case of molding a study around trying to substantiate a predetermined (and financially motivated) goal.

Unfit to do what specific "software development" job? There are Indians working this very vast field. Presenting them with a test they fail does not change that fact.

If your study results in a claim that contradicts reality, you are at best doing it horribly wrong and have at worst terrible ethics.

  36,000 engineering students from IT related branches took [the test]
  over 2/3 could not even write code that compiles
The information missing from the report is:

1. What stage in their IT education were they in?

2. Was the test given in a language the students were meant to be familiar with?

I've been programming professionally for over 15 years, yet I wouldn't even be able to write "Hello world" in Haskell without having to look up a tutorial - because I've never used Haskell.

I don't believe this. I have worked with many offshore teams in the last 15 years, and the teams in India have been roughly on par with teams in Eastern Europe. The only advantage of European teams is that they seem to be better educated in areas outside of software engineering, like mathematics.

I'm an Indian, and the article is factually correct in that most engineers are unfit for software development albeit I wouldn't quote an exact figure.

Perhaps your experience was such because most of these "engineers" remain unemployed, take up another profession or join the Indian sweatshops (ie, TCS, Wipro, HCL, Infosys, Tech Mahindra, Accenture etc..).

This would explain the poor code quality that often comes from off-shoring.

More and more software companies purely based from India deny this claim.

One of many examples competing (successfully) with (for example) California companies (who pay around 2..4x their devs): http://enpass.io

Disclaimer: I'm not Indian but European and not enpass owner/user/developer.

5% of all engineers in India is still a sizable amount.

If they pay 2-4x the average salary, it's not surprising that they get the 5% top of developers, no?

In any case, this isn't an India-specific problem; the fact that most candidates to a programming job can't program has been reported in various countries, which is why Fizzbuzz is still a thing.

This is not good journalism. Facts and figures about the study are missing and the title is absolutely click-bait inspired.

Meh. "write code that compiles" means to me the ability to do zero defect keyboarding. That's not so important - being able to interpret the compiler output to say - Oh, I typoed that variable name/I missed that semicolon/etc. - that is what matters.

For me, the more problematic evidence comes from the interaction colleagues and I have with actual Indian engineers. Westerners are far from perfect, but I've had a lower success rate with Indian outsourcers. As I understand it, Indian cost of living is at most 1/2 that of western countries, so I would expect a 1/2 price differential for the same quality. But that doesn't seem to happen. I'd guess it's due to upper management failing to see the difference between the Indian quality at 1/2 price and Indian quality at 1/10 price, and choosing the latter, making the former scarce.

I think a lot SWE in India might not even like writing Software. It mainly provides a way to get out of the country or at least make a decent living. A lot of these Outsourcing companies will hire pretty much any College grad and just teach them how to program.

I know many really great Indnian developers. I know even more really bad Indnian developers.

I have an impression that there is huge social pressure in India to become a developer. What makes sense is because it is a good profession.

Because many of these developers have too little capacities to become a programmer, hence, there may be a stereotype of poor programmers from India.

It would be interesting to see at least the full article. Anyway, even if this information is true, the 'study' is useless if not compared with other countries.

I doubt their claim without stats and facts, Company falsely claiming % for promoting product

1. have they reached 100% students? to claim 95%? no they have not reached there are 500+ colleges in a single state 2. From which engineering year student given test, for first year most of the colleges have same syllabus for first year 3. Company itself mentioned using automata it was found that the machine learning score was able to predict 22.6% good candidates

I'm curious to take a look at the questions in that test. Would be interesting to see how well they are aligned with the reality of software development. Anybody got some of those questions?

I'm confused. Is this saying they couldn't write code for machine learning, or that the evaluation program used machine learning.

I think it's this: http://www.aspiringminds.com/technology/automata

"Automata is world's most advanced and only programming assessment that uses machine learning for grading programs."

"It can be combined with our technology specific modules such as those in Websphere, Power Builder and Java Swings, to form a powerful suite to test candidates for different positions."

That pretty much sums up anything you might want to know about the nature of the test.

Probably the latter

Has anybody here taken the automata test? Thoughts?

Disgustingly irresponsible reporting. All this can do is encourage racial discrimination in hiring.

The original article is from an Indian news site targeted at Indian business people. This is relevant to that audience if they are looking to hire local developers. The study is, not coincidentally, put out by an Indian company whose products are employee tests and certifications.

I'm not sure where race/national origin factors into this if all the participants are Indian. Perhaps you could point that out? I am vaguely aware of India's caste system but I don't see any mention of it in the original article.

Very annoying clickbait title. Article mentions engineering students only. "Over 36,000 engineering students from IT related branches of over 500 colleges took Automata ā€” a Machine Learning based assessment of software development skills ā€” and over 2/3 could not even write code that compiles."

It talks about college grads!

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact