Slashdot Mirror
95% Engineers in India Unfit For Software Development Jobs: Report (gadgetsnow.com)
An anonymous reader shares a report: Talent shortage is acute in the IT and data science ecosystem in India with a survey claiming that 95 percent of engineers in the country are not fit to take up software development jobs. According to a study by employability assessment company Aspiring Minds, only 4.77 percent candidates can write the correct logic for a programme -- a minimum requirement for any programming job. Over 36,000 engineering students form IT related branches of over 500 colleges took Automata -- a Machine Learning based assessment of software development skills -- and over 2/3 could not even write code that compiles.
...that we can discuss the abysmal skills of your average Indian IT worker, without being branded a racist, or using excessive PC language.
Completely validates that report. When my last employer decided to fire the American citizens (forcing them to train their "offshore" replacements in order to receive any severance) that built the products and systems that made the company a success, those of us that remained discovered that we had to rewrite everything they produced (with a much smaller staff, of course). The greed of executive management results in far worse products for the customer - but they got their bonuses, so they do not care.
What the numbers would look like in the US.
Pretty sure my parent company still outsources to all of them. I hate making large broad statements, but I've never yet met one I was impressed by. Seems to whole business model for outsourcing revolves around everything being so cheap you can rebuild it 5x and still come out ahead on direct project costs. As for impacting the business with garbage software, that doesn't cost anything, right?
I heard that 3/4 of the people working on Windows 10 couldn't write code that compiles, so I understand why they are hiring from India. ;)
Anons need not reply. Questions end with a question mark.
I would say in a whole, true software engineering has been completely watered down and very disappointing over the last 10-15 years. From all the way down in school systems with STEM and all they way up with these 3-4 day crash-course 'bootcamps' and seem to manufacture quick hot-on-resume-paper skills without experience is really the problem. And even on top of that, how many people just 'google' their way into a job or solution? No one thinks anymore, we are in an age of just-give-me-the-stuff mentality. Don't care how or why, just blindly take the answer and move on. You don't grow as a competent and efficient engineer that way.
Coupled with the fact that any business, company or dev shop wants talent in our psychotic digital age, this reminds me nothing more than a massive amount of people doing nothing more than to try to get their foot into a hot job market and doing nothing more than trying to flip a huge salary for 6-12 months. And that's why I say it has very little to do with India.
I am shocked! I cannot describe how shocked I am.
Why would Engineers write code? Shouldn't those Engineers get back to driving the trains and leave the programming for the programmers?
/ Call me a Software Developer. Call me a Programmer. Call me a Code Monkey even. I am not an Engineer. Calling programmers "Engineers" is stupid. It's like calling janitors "sanitation experts" or secretaries "office administrators". Call a rose a rose and stop all this silly flowery job titles.
"That's the way to do it" - Punch
yeah either that or the test was bullshit
~.~
I'm a peripheral visionary.
The travelling salesman problem?
Slashdot, fix the reply notifications... You won't get away with it...
And I'm an American with several decades of experience. But, I'm not worried because my compiler kindly points out my mistakes. Good thing there aren't any reporters around to tell the world that I left out a semi-colon here, misspelled the name of a class there, or forgot to remove one of the arguments to one of my own functions whose API I'd just modified.
Most of them simply have no proper logical reasoning skills. Even if they do, they are likely to have sub par English skills which makes the issues they write still seem like nonsense . Yet they manage to win contract after contract, simply because they are cheap
From what I've seen in my grad program (mid-tier US university), I'd say the figure is closer to 99%.
They're expert liars, though.
There's a difference between "training" and "provide fundamental education".
I think there are many talented and smart developers in India (as anywhere else). The biggest issue is that they mostly want to work for very large companies (prestige), they are in a hurry to be promoted to managers (many are not good at managing anything but it's all about the title) and thus good developers become weak managers. This depletes the software developer pool so they have to hire people less and less qualified to do the coding.
Another is that there are a lot of "software consulting" companies that handle outsourced work, they tend to have some good developers and a lot of "junior" developers, so when they sell themselves to a customer they can say they have a staff of 100 developers ready to go. This is compounded with the problem of developers trying to get promoted into management (again, title and status are very important to people).
I am not sure if 95% is an accurate number (seems a bit high), but the problem exists nevertheless.
I have read that a lot has to do with sociological issue of being used to a caste system, and while it's not as prevalent as it used to be, rank and status are very important. While this is also true in many other countries (I have worked with many Eastern European and Far East companies), India remains as the place where every developer seems to be looking for a promotion. Some companies placate the developers by giving them over-inflated titles like chief architect or senior staff engineer; but in a company with dozens of chief architects the title no longer has a significant meaning.
Anecdotal evidence: I worked with a developer who was young and his mom kept emailing him to get promoted to a manager so that when she went looking for a wife she could pick from a nicer "deck" because he was a manager ( a deck of pictures/bios is how moms and matchmakers and astrologists get together to determine who gets to marry whom, it's very complicated from what I have seen). I thought it was funny, but he was very serious that the "quality" of a wife his mom could get depended a lot on where he worked and what his title was. At one point he lobbied to get a temporary title and we put him on a short term support project where he was handling issues for one single customer and had a temporary title of a "Senior Customer Manager". He was married within 3 months.
It's a company trying to sell their assessment products that are more marketable the higher the number they manage to produce out of their "study". Extrapolating "36,000 engineering students from IT related branches of over 500 colleges" to " engineers in the country" seems a little generous as well. Most of the students in IT related branches I've met are also really crap at programming - because they aren't actually doing programming or because they are first years who haven't managed to learn anything yet.
That said most of the people I have interviewed for programming positions I would put in the "can't program" category too. Not 95%, but probably 60%.
And I would expect the Indian IT education system to have more than its fair share of really bad "colleges" compared with say the US (and note that the US has things like "ITT Technical Institute"). It's a bigger country population wise with worse infrastructure and government oversight. The good programmers seem far more likely to go and get a job overseas than they do to take up an academic career in an Indian college...
I have worked on many outsourced projects. So much so, that my position transitioned from being a software developer to one who provides development support. So I do the things they can't complete. Anything from browser interaction problems to performance to security. One might think I have a jaded view - and this is something I am always assuming that I have. I have seen everything from absolute incompetence to some "diamonds in the rough".
That said, I believe the issue in India is the way the problem is approached. Rather than let the gifted students percolate out of the system (a focus on quality), they encourage everyone to enroll and encourage the institution to graduate everyone. I can't comment on the quality of the education, but I suspect it spans everywhere from decent to criminal. The institutions are not schools - they are factories.
But I still think this report is BS. Perhaps I will change my opinion when my job is transitioned to India. This is in process right now.
Is this just an Ad for Automata? Because that's what it looks like.
Recently for the first time I had experienced working with software developers from India. They were all recent migrants working with a consulting company. In my project team we had about 20 of these engineers that I had to manage and for the most part they were pretty good. On the plus side, they were hard working and keen to learn and best of all they were able to LISTEN and take responsibility on what was sometimes quite a stressful project. The negative side would be perhaps having the courage to take initiative and move the team in a new and better direction, but maybe that will come as experience grows.
Overall a pretty good experience. I would definitely work with some of them again.
Though I'd love to believe this is true, promising something you want to believe is the easiest marketing scam of all.
I've worked enough with Indian developers to know that although the percentage of incompetents is high, it is not close to 95%
Automata, the tool used for this, is a commercial job interview assessment tool.
This company benefits greatly from making it appear that most hiring candidates are unfit for the job; it creates a need for their product.
Slashdot social media options: AIM, ICQ, Yahoo, Jabber and Mobile Text. Why no MySpace?
So there is probably a lot of truth in the reporting, but the shock value of the story comes from the numbers. 95% you say! Oh my! We cannot have any Indians write code! The details, in this case, matter a great deal, so lets take a look at some of the unanswered questions that may impact the accuracy of that number.
* What does "...not write code that compiles" mean? Were the people being tested provided an IDE? I'm an expert Java programmer, but if I were to open up a text file and type Java code, odds are pretty good that my code won't compile on the first try. That's what IDE's are there for - to fix the inane syntax issues. But lets say that the IDE's were provided. What sort of languages were used in the test? Were the test takers familiar in the language being used? Was the measurement really meaning that they ran out of time to make the program compile or that they were incapable of making it compile because they really weren't a programmer? I note that the "cannot even compile" statistic is 2/3 - not 95% according to TFA. Still bad, but details are needed to see what was being measured.
* What does the sample mean? TFA says that the sample size was 36000, but how does this compare to the universe out there, and who made up the sample? Were these graduates in computer science or first year students or people already working in the field? What was the level of quality for these universities? Where did the 5% who did good come from, and did those 5% come from the really good schools? Was the sample size structured to represent the real world distribution of quality in educational institutions?
* Bias: who is aspiring minds, and what is their motivation? Are they tied to a particular agenda? Is there a competing country that wants their programmers to be hired over Indian programmers pushing these stats? I will point out that there were numerous doctors pushing the agenda of the tobacco industry, and numerous scientists pushing the agenda of the oil industry (global warming). So, yes, the affiliations need to be clear.
I will also point out that in the silicon valley, Indian engineers are present in high numbers. And a lot of the clamor for getting Indians into the US comes from companies in that area. If 95% of them were useless, I can't help but think that there would be less demand.
There is no such thing as luck. Luck is nothing but an absence of bad luck.
95% percent of Software Development employers unwilling to train people into the job.
Programming isn't an assembly line, where you can train people in the steps and they just repeat it. You have to have an ability to think, and come up with new solutions to new challenges. And when faced with alternatives, be able to pick the one that's the best for the task at hand, and understand and be able to explain why.
Training means very little. Someone who requires training is near worthless as a programmer.
There's a difference between "training" and "provide fundamental education".
Indeed. Education is meant to steer a person into being able to find solutions, and a thirst for more knowledge.
So far the company work for have had multiple interactions with outsourced 'developers'. I can't even use the word "incompetent" because that implies that they have at least some skill at their job, and that's patently untrue.
Every. Single. Project. that they were involved in became a nightmare. The time and effort required for babysitting them, and correcting their (sometimes incredible) mistakes was greater than our own work.
While I'm sure there are exceptions, in general I would say that this article is completely correct. I wouldn't trust these people to flip my hamburger correctly let alone operate something that uses electricity.
Soooo, a "machine learning" system made this evaluation. I'll bet other machines got perfect scores. Bias much?
I work for a university here in the US and have an opinion about why this seems reasonable. It is a sociological problem like someone here already mentioned. I was one of the few developers on my team back in India who really knew their shit when I worked back home. I think this stems from how I got into my programming career. I wasn't trained in programming by my company or college. I graduated in Electronics and picked up programming to make games and got good at it because I liked it. A lot of people who work in IT in India just simply shouldn't be working in IT - Not because they are not capable enough (It is, after all, a skill that you can learn with time) but because they are just not interested in their career. Careers in India, if you want to earn good money, are very limited. IT is a field that gives you the best investment to return ratio(4 years of college and a cushy job at an IT firm). Familial pressures encourage young men and women who have no interest in the technical side of things to pick up careers in IT because they can pay well in a short time. This leads to a lot of people who would frankly be better off in other people-oriented careers to slog in something they have no interest in. It is also a personality issue. This may seem like generalizing but I have experienced it. I tend to get along well on a personal level with the people working with me here in the US. Call it the "birds of a feather" phenomenon. Working in computer science, you tend to pick up a certain type of personality and social skills. I like games. My coworkers like games. I like talking about technology and memes. My coworkers like talking about technology and memes. I find I get along better with other "nerd" types. Working in India, even in college was a nightmare because I shared my time with people who are just a different breed. People who fit other social archetypes that I don't tend to get along with. I didn't understand the humor, wasn't interested in their discussions and felt that most of them were generally people I wouldn't get along with. So people who are just not wanting to work in IT and just want to be managers end up spending years in this career. It is frustrating and pushes off the kind of people who should be programming. Just my two cents
Best Programmer I ever knew was Indian.
Also, most of the worst programmers I ever knew were too. Just like everywhere, they produce quality and low quality programmers.
"That's the way to do it" - Punch
Then they should exercise more!
Unfortunately I am not surprised. But it's not an India specific problem. I interview people from all over and end up rejecting very high percentages of them.
Where they come from (school, country, degree level, etc) has so little to do with how well people do that I just ignore all that. You can either think a problem through and code up a solution or you can't.
I would really love to take their test. Knowing what they ask and how well I got graded on it would definitely help me judge if this test has any chance of accurately measuring skill.
My supervisor at Cisco had me sat in on a conference call with the developers in India trying to convince them that they didn't fix a crash bug and incrementing the build number by one didn't fix the problem. The developers tried to get me involved to go against my supervisor. I pretended to have phone problems and played dumb.
You seem to confuse training with education.
Yes, you need to acquire basic computer knowledge and understanding.
But on the job is not the time for this.
Code is a tool. Engineers know how to use tools.
But don't expect to learn to code in Engineering school. You should have gotten that down in high school or before. I've never known a computer or electrical engineer that couldn't code at all. Some haven't in years, aren't great at it, prefer assembler etc, but they know how.
Which doesn't make them 'coders' any more than a mechanical engineer that touches a lathe becomes a 'machinist'.
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
Who needs correct logic?
Cut-and-paste, baby!
Other key skills include begging others to do your work in support forums.
Different but similar study by the same company it looks like they are testing engineers from all fields just so they can publish an alarmist report conveniently failing to mention that it's just SHOCKING that a mechanical engineer wouldn't know Object Oriented programming! /s These guys are selling the very test they studied. This is spam that got picked up because it looks legit and regular news (and slashdot editors) don't bother to investigate when given a chance to print outrage-pr0n. It's scaring students, universities, and employers into licensing the test or test-training resources.
. . . . I had to deal with outsourced services to India, back around the turn of the century, when I was working for a Dot-com. We were experimenting with outsourcing. The boss had contracted some Indian services firm to take 1000 Material Safety Data Sheets, in scanned PDFs, and type the contents of each into a Microsoft Word file. As I recall, he paid 500 bucks for it.
Average results were 20+ errors per page. Spelling, words missing, whatever, it was there in the results file.
This compares to a similar contract, same task, somewhere in Flyover Country. Zero errors, and notes pointing out mis-spellings in the ORIGINAL PDFs. Cost 5 grand, but worth it. . . .
"only 4.77 percent candidates can write the correct logic for a programme -- a minimum requirement for any programming job."
I thought, knowing that computer programs are never called programme, neither UK English, nor US or Australian English is different in that way.
http://www.dailywritingtips.co...
I looked into it and based on another study they did yeah, they tested all engineers. It looks like they wanted scary numbers so employers would be alarmed into buying licenses to administer the test, and universities would be alarmed into buying licenses and materials to teach the test. Because this looks like a seriously fudged research-design.
You don't employ a cabinet maker / master carpenter to frame your house. You get one to supervise the work of others. You only need guys with basic skills to nail 2x4s together every 14". Similarly you have the more expensive talented guys do the finish work because you want the miters on your moldings to match up perfectly, your doors plumb so that don't fall open or closed or stick, your drawers to slide easily, etc.
The problem is that you needed guys to nail 2x4s together every 16". But your idiots nailed them together every 14" and now the whole project has ground to a halt because you ran out of 2x4s.
Later your project will slow to a crawl again because the guys nailing the sheathing, roofing and sub-floors have to eyeball each joist and stud as the panels aren't marked with a 14" nailing schedule. Worse, each panel will need to be cut to a multiple of 14" in order to support the ends, with the maximum panel length now being 84" instead of 96". You've just added about 12.5% to the material cost of all those panels, and a bunch of additional labor cost too. Disposing of all those 12"x48" offcuts isn't free either. Don't forget to buy 12.5% more nails.
What are you going to do about the windows? They were ordered as standard windows so are 30" wide to fit in the 30.5" gap across two stud bays, leaving .5" for shims to plumb them up. But for some reason those gaps are only 26.5". Will you return the standard windows, eat a 15% restocking fee and wait a few weeks for custom 26" wide windows to be made? What does that do to the aesthetics that your customer cares about? (Narrow windows suck.) Your other choice is to rework the studs for each window to allow for the standard windows anyway. More 2x4s needed, more time and head-scratching. BTW, don't forget that the building inspector is going to need to sign off on those mods; he stamped the original design at a standard 16" OC but you have deviated from those construction plans, and he needs to ensure that the actual construction is to code. More delay as you need to draw up revised construction plans for approval.
That 12.5% cost overrun keeps coming back to bite. Insulation typically comes in 14.5" widths to fit in one 16" OC bay. Your bays are 12.5" wide and insulation doesn't come in that width. Every roll or slab of insulation needs to be cut to size because overstuffed insulation doesn't perform to its rating and the building inspector will insist. You now have to buy 12.5% more insulation than you'd have needed if you'd nailed at 16" OC. As an additional bonus what should be a simple 1-2 day insulation job now takes several times that and one way or another you're paying those salaries. More disposal to pay for.
Eventually you get to drywall or plastering. Back to the 12.5% material cost overruns you go on boards and screws. Back to the landfill with 12"x48" waste offcuts.
I get it; in your head it all sounded so simple to justify laying off the skilled workers and replacing them with drones. Sure your drones can raise their game and put a house together. How hard could it be? It's just nailing, right? The problem is that drones cannot prevent idiots in management who apparently don't have the first clue about construction from totally screwing up the project. Your skilled workers would have pointed out all the problems with your 14" stud nailing schedule before a stud was cut. Your drones just went to work and killed your profit doing what you told them too. Your framing guys really enjoyed getting chewed out by the roofers, insulation guys and plasters as each of them in turn realized what a pig of a job this house is. Nobody was really happy when you fired the skill workers before, but now you really have a staff retention problem.
Thanks for providing a great analogy.
Final sting - Your customer is pissed too. Not only did he need to push off his moving date because of the delays, but heating and cooling costs are about 10% higher than they should be because about 12.5% of the R38 insulation has been replaced by 10" of R1 wood. He's going to a competitor for his next house.
The current trend in software is to distribute a problem to a bunch of commodity servers. That is a good strategy for extremely large data sets, but it can also be a crutch to compensate for bad algorithms and inefficient code. Rather than focus on getting the algorithm right and solve the problem using 80% fewer instructions; you just get more cheap computers (and electricity, and heat, and...) involved. I wonder how much this strategy has spilled over into the headcount equation. Don't get one great employee that can solve the problem by himself on a $150K salary...instead get 10 programmers that can try to do the same thing collectively at $15K each. It might look the same on the corporate balance sheet, but almost guaranteed that the solution by the one qualified engineer will be much better than the code cobbled together by the team of 10 mediocre or incompetent programmers.
First this is a claim about a proprietary test from a company earning money from it, information about what it actually does is not available without paying AFAIK.
Also I've read here and elsewhere how new employees can't code no matter where they come from or graduated. Some even claim that almost everyone interviewed didn't even have basic skills.
A moron that earns money by selling the test the article is about. But I agree with you, even skilled programmers make trivial errors that a compiler will complain about - in the real world they fix it in a few seconds, depending on how the test is designed that may not be possible.
I actually did wonder if they are talking about software engineers (who should be able to code - but maybe not in the test language without learning it first) or computer engineers (who really can't be expected to be programmers). Bad article with little information.
Firstly, is their work ethics...lazy and too much elitism and bureaucracy.
Then comes their limited command of English; and the English mistakes, that come either due to laziness and lack of practice.
They are also yes man by nature, and do not know when to say no...ask them something impossible, and they will say yes.
They also underestimated their projects.
They also have a mysterious gift of over-complicating the simplest of the problems.
Furthermore, they twist the truth more than we are used to.
Most of their CVs are fabricated; their knowledge and experience is sub-par.
They also exaggerate the actual experiences they had in the past.
And they do not respect women...
Finally, to hit home: if I pay big bucks, and you put me talking not with an actual local expert, but with an Indian guy, why should I pay big bucks, and why should not I cut the middle man, and pay them rupees?
You seem to be making the mistake recruiters are these days and assuming that if someone isn't a carbon copy of a current employee (good luck finding that) they're unhireable. I'd argue this erroneous assumption is the cause of most unemployment today.
No, I make the mistake of believing that there are people out there that can figure things out without having all their food chewed for them. Too often i get disappointed.
I have the ability to think and come up with new solutions to new challenges. If I'm suddenly required to program in a language I don't know, I'm going to need training of some sort. I can learn it myself, but it's probably going to be more efficient to find me some training, and my time doesn't come cheap.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
I agree with your point regarding training vs education. However very little development is green fields. New developers should receive product training in your domain specific field. You do have to introduce people to the existing product and the thinking that went into solving a specific problem. Otherwise you wont leverage the value you already have, and your code-base will quickly evolve into a fragmented set of different solutions. Programmers need to know the fundamentals and the inherent ability to think and come up with new ideas (education and skill), but they should also learn and take advantage of things other built before them (product training).