American firms learned that once they have thousands of resources in India it's more cost effective to just hire them as direct employees instead of going through an outsourcing company. In the long run this probably works out better for everyone, except the owners of the outsourcing companies.
>it's more cost effective to just hire them as direct employees
it also makes it easier to bring them later (after 1 year) into US on L1 without all these limitation of H1 like prevailing wage, yearly cap, spouse of L1 can work, GC seems to be easier (at least that seemed to be the case 10 years ago when I last paid attention to immigration)
Well german firms aren't there yet in my opinion. They propagate offshoring as the opportunity to lower costs by 90% while preserving the same quality. In my company the last "transformation" talk made clear that no department will be left untouched in the next off and near shoring activities..
Which german company? That sounds stupid. I think no company profited from that over the long term, ever. Management must be stupid to think 90% of costs can be cut... if anything, costs may go up
I am not that old, but I remember reading a headline like this on Slashdot once... The results have been less than stellar. When my non-tech enabled friends complain about tech support, it's usually with a fake Indian accent. I think the "IT professional" in India has hurt quite a few off-shoring companies.
I suppose it was a long time coming. Why send your work half way across the world, for it to be done/corrected 3 times over with often poor results, when you can do it once and do it right onshore. In my experience, offshoring is the absolute epitome of a false economy!
The type of company that hires these companies don't have the corporate structures necessary to build decent software either. They'll build crap 3 times over with poor results and cost more doing it.
I'm sure you've heard about those people with 10 years of experience that are unable to code fizzbuzz? They hide out in those large corporate environments.
I'm pretty sure the parent posters were referring to large corporations whose core product is not technology as such, although even there, you'd be surprised (or maybe you wouldn't). You don't have to get far from MS/Apple/Google/FB/etc. to walk into an alternate universe of complete mediocrity, where engineering and development are viewed as cost centres and/or fancy forms of typing...
Even within those companies lurk a lot of mediocrity, either acquired or homegrown. MS and Google have had some stinkers over the years, apple seems better, but the probably keep the worst of it internal.
If you ever read stories about internal political wars at those places, you will realize they aren't so great to work there.
Specially taking into account the type of assemments being done to the employees, or which R&D unit is driving the main products (e.g. WinDev vs DevTools).
Yes, places like Microsoft System Research are great, but that is just a unit from the whole corporation. Same applies to the others.
Because the majority of the customers are not IT companies, they just want a bunch of developers to help them provide a customized solution to a specific problem.
After that they are perfectly fine not having IT people around.
The time each company had a set of developers to sit around and help them on as needed basis, is gone and won't be coming back.
Those that want to keep around onsite development need to upskill their social skills, and also be able to bridge between developer, architect and business relations.
> After that they are perfectly fine not having IT people around.
I've never seen this work in practice. It always takes almost as many people to properly maintain a solution as it did to build it, there is no "done" state where the can get rid of the developers. It was accepted as far back as the mythical man month that maintenance is where the vast bulk of software costs are.
> Those that want to keep around onsite development need to upskill their social skills, and also be able to bridge between developer, architect and business relations.
At this rate we're about a decade a way from expecting developers to be one man corporations.
> I've never seen this work in practice. It always takes almost as many people to properly maintain a solution as it did to build it, there is no "done" state where the can get rid of the developers. It was accepted as far back as the mythical man month that maintenance is where the vast bulk of software costs are.
Those companies see software development as a cost and not something that need to be kept around.
Usually we also don't have someone sitting on our building to repair the doors when they get broken, we call someone to repair or replace them, when the need arises.
> At this rate we're about a decade a way from expecting developers to be one man corporations.
If you are living in an European country this has already happen long time ago.
In countries like Portugal, many companies are hiring developers on "Green receipts", which are literally one man corporations.
Similarly in countries like Germany and Switzerland, many consultants are freelancers, or work for a consulting company that plays the same role as agents do for sport and movie industries, giving them a fee. Also expected to be an one man corporation from the customer point of view.
The glorified idea of the developer that codes all day long on his cubicle/desk without anything else to worry about is long dead, at least in western part of the globe, when targeting the industries that don't sell software.
It always takes almost as many people to properly maintain a solution as it did to build it, there is no "done" state where the can get rid of the developers. It was accepted as far back as the mythical man month that maintenance is where the vast bulk of software costs are.
... accepted by companies run by people who read the Mythical Man Month.
Then there's the rest of the corporate world. These things you say are absolutely true, and yet not a day goes by that a company isn't learning that lesson, quite expensively--or better yet, failing to learn that lesson over and over.
Just take every comments like these with a pinch of salt, some people are so against Indian IT firms that they spew all kinds of nonsense. "automation and AI" being one of it.
I already reviewed code delivered by several European consulting companies with worse quality than many Indian IT firms, including those on the article.
The sad reality is that the way software is delivered at enterprise level, there is hardly any company that can be shown as example of excellence.
Only price and the output of applications matter, not how they are running inside.
it was a direct answer to you saying 'they must be doing something right' the implication being that while they are able to find paying customers, what they are /not/ doing right is the actual work of developing software.
the parent points out that this market inefficiency exists because the people making purchasing decisions are not technically competent enough to truly evaluate the vendors that are selling them these development contracts.
A company can make money selling a bad product, it happens all the time; and its a well know phrase in finance that 'markets can remain irrational longer than you can remain solvent'
First of all if you going to quote Keynes, use it in the right context. We are not talking about stock market here.
So you are telling me that these Indian outsourcing companies who have been growing for 30 years with billions of dollars in revenue is all because they are selling a bad product.
So the Indian companies are like wolves eating feeble Western companies like lambs. Do you realize how ridiculous that sounds.
Sadly the article does not specify whats automated and slashing these jobs. I suppose the majority of these terminated positions are service roles, like help desks / call center?
Not sure about the quality of this piece. One company, Cognizant, is obviously in some form of trouble. The others are laying off a few hundreds/thousands jobs, on a base of 100K+ employees, as part of the performance review process. There's also no mention of hiring being reduced for any of these companies.
Most of us in the Indian IT industry, have probably seen the writing on the wall for about close to 6-8 years now. One way Indians can handle this adversity is to switch from working for service companies to working for firms that build product.The silver lining, although a very tenuous one, is that venture capital and private equity inflows into India in the areas of e-commerce, digital marketing, fintech and software automation have been steadily increasing over the last decade (although not in 2015-2016.) That provides opportunities for experienced professionals -- the cohort that seems most affected. Those with the experience of managing people, technology and operations may find the downturn easier to navigate.
Could be. There is nothing sacrosanct about Indian workers. (Or any others for that matter) The reason I was advocating that folks with experience and talent seek out product companies, is that as a group, they are less sensitive to worker wages. A talented programmer, product manager, marketing executive or ops person will have at least an order of magnitude greater impact on the company's bottom (or top line) than they would in say a pure body shopping IT outsourcing company. So as an entrepreneur, for the right candidate, I have no problem paying 2x the market rate. A good employee will more than make up for the extra cost. In the outsourcing business though, the business model is very sensitive to labor costs. Not so with product firms.
> which is moving towards increasing automation, use of artificial intelligence
It is such baseless stuff which irks me every time, I have read from MANY papers that automation is doing this that without explicitly saying X automation firm screwed up the revenues.
I would like to see visible proof that "automation" is eating into the revenues, no report says that, they just say "automation and AI"
The journalists don't know difference between AI vs ML and everyone is acting as if "AI and automation" will destroy X firm, no, they won't. Anyone who works in the sw field would know that there are many manual things done in the project so much so that there wouldn't be an "automation firm" which would sell blah blah robots to do human work. It is way off into the future, if it'll ever happen. This madness has to stop.
I did not mean the people working in IT, I meant the journalists, they just say "AI is eating the world" without knowing anything. "automation and AI" are eating into the revenues. Why exactly is building these "automation" so much that the revenues are getting affected?
Maybe these crappy IT consulting firms make a large amount of their income from selling the 'same fix' over and over again. When the customer moves to a newer system the number of legacy problems drops dramatically, and the new systems can scale far beyond their current needs before capacity problems increase billable hours again.
I've seen it myself doing MSP work. Crappy old system breaks with every program update, every windows update, the wind blows the wrong direction, etc, etc. Client updates to new computers with SSD, new software based partially/mostly in the cloud. Full and correct integration with active directory. Their complaint load can easily drop 5 to 10 fold, mostly from less legacy software issues.
Why are you so convinced that it's not happening ?
Each time something, which used to be done by humans, is done by a machine, that time spent doing that something is no longer payed to a human and only a fraction goes to the makers of the machine - that's the whole point of automation - reduce costs and increase productivity.
It doesn't lead to [more] humans being payed more money.
I think intuitively it makes sense, but coming up with undeniable proof is quite hard in the globalized economy - the effects of automation often happen further away in the supply-chain graph.
The 'modernized' company just stops importing or paying for X, because they can now do it 'in-house' for a fraction of what it cost them before.
The effect of that is that someone somewhere on the other side of the planet might not be able to find a job doing X as easily as they did before. Or their company goes bankrupt. Is it because of automation ? Could be.
It might create new opportunities, but those require newer and 'higher-level' skills - and as these opportunities arise, someone somewhere is already hacking away trying to automate them. And we're getting better and better at this, so the new opportunities are only available for a short period of time, until someone comes up with a brilliant new way of doing things. The software world is a great example.
I personally think it's very real and being in denial the way that many people in SV are, does not improve matters. We (the technologists) should at least ponder upon these questions and figure out solutions - otherwise we're leaving that in the hands of politicians, which serves as zero comfort to me.
I know that in some or the other sense tooling is happening, efficient tools are written. For instance, I wrote a python script which saved $10k per year, this doesn't mean that companies are going bankrupt because of "automation and AI". This "automation and AI" wave in newspapers is a figment of their infertile imagination.
I think the role of AI in eliminating these jobs is overstated compared to the role of Cloud, Dev Ops automation and faster iteration cycles afforded by relatively better abstractions and tools for building software compared to 10-15 years ago.
A lot of these large IT companies relied on long term maintenance contracts for infrastructure and legacy software with unnecessarily large teams on self serving time and material (T&M) models. But these days, even large enterprises have to innovate to stay in the game and their experience with top end Agile consulting firms made them realize that you can do with much fewer people with automation and hiring better. On the other hand, the large IT companies got stuck with their old model and without training and experience on challenging projects, the talent atrophied even while having to eek higher salaries every year due to inflation and talent loss to startups and others.
However, the numbers being reported so far aren't huge compared to the employee sizes of these companies. For e.g. 5% layoff would actually be a reasonable trim on an annual basis. In that sense, the carnage hasn't started yet.
Yes, but automation alone isn't causing layoffs! Bad management is. They are just hiding their own failures on the pretext of "risk of automation and AI", it is nonsense!
If you've ever spun up a VM rather than unboxing and configuring a physical server, you have the visible proof that you seek. If you've ever used Chef/Puppet/Ansible to configure X number of machines rather than configuring each one by hand, you have the visible proof you seek. And that's before getting in to the recommendations and optimizations that AI/ML can offer... it's just the old fashioned AI/ML, MBA's with a spreadsheet, deciding they don't need all those unboxers/configurers anymore.
This thread is not about web hosting companies. Chef did not replace jobs in the Indian IT sector, it just made it easier to deploy things.
We need to separate our discussion points.
I know that tooling is improving considerably, and rightly so, but that doesn't mean that IT companies will now be downsizing to 10 because of chef or puppet.
You're saying that it isn't reasonable to imagine that companies might decide to do their own virtual infrastructure rather than hiring Indian IT firms to build infrastructure for them? I will keep my own counsel on that.
I am not saying anything of that sort, all I am saying is that layoffs are not happening because of "automation and AI".
If a company wants an inhouse dev team they'd do so without "automation and AI", the thing is, they don't and that's why enterprise software companies are formed.
Sure, I can bake my own bread, grow my own vegetables, mine coal for electricity myself, but I don't. I do what I do best and leave the other stuff to others. It is upto companies to decide if they want to build infra for them.
Agreed. I'm not going to pretend or claim I understand the market forces leading to these firms downsizing, but I too find the idea of tooling like Chef and Puppet causing a measurable downturn in staff numbers at the big Indian IT firms hard to believe.
Whilst DevOps automation tools like Chef/Puppet/Ansible/Whatever have undoubtedly increased in popularity, very often they are just used as nicer ways to achieve the same thing IT professionals have been doing forever in shell scripts or similar.
This is purely speculation, but I wouldn't at all be surprised if US companies are simply less interested in hiring Indian outsource firms given the current climate and uncertainty around the future of work visas in the US. It's not hard to imagine such companies indicating they will spend less with these firms for the foreseeable future. I don't think anyone believes eventual reform of the H1B visa, the lifeblood of Indian IT firms in their biggest market, is going to be good for the Wipros of the world.
> Whilst DevOps automation tools like Chef/Puppet/Ansible/Whatever have undoubtedly increased in popularity, very often they are just used as nicer ways to achieve the same thing IT professionals have been doing forever in shell scripts or similar.
Exactly! The mgmt of the above stated firms sucks, they had no vision or a sense of direction, they make bad decisions and when they are asked questions like Sikka (infosys CEO was asked) Why do you earn millions per year and why do you take 50+% hike and still are unable to deliver then the hide behind "automation and AI" :)
it is mediocre CEOs and C*Os and a lot of other folks which are causing this downturn plus investors too, Google "eliott management letter" to get an indepth real reason behind this wave of "layoffs due to automation and AI".
> Whilst DevOps automation tools like Chef/Puppet/Ansible/Whatever have undoubtedly increased in popularity, very often they are just used as nicer ways to achieve the same thing IT professionals have been doing forever in shell scripts or similar.
You'd be surprised how many sys admins and especially developers have never written a shell script.
One of our code tests is to write a simple grep clone, the results seem to indicate that none have ever used a command line. Nearly every submission seemed to think that you could only pass one "-x" (x the placeholder) switch to a command.
This could absolutely be true, but in the context of this discussion I find it hard to imagine someone using dev ops tools like these without having written a shell script in their lives - even the most basic use of these tools is probably going to template a shell script at some point. I'd be shocked if the shell command resource provider syntax in these tools coupled with their templating engines is not by some margin the most commonly used feature these things offer.
Having worked for one of those names, I have to guess that this is probably clearing the bench rather than removing employees with billings.
More often than not, the bench exists to promise a customer that they have the people to "start yesterday" on the project.
The trouble with choking off the supply into the industry is that often the demand dies out as a result rather than increasing the price of a service - complete rewrites get shelved, new development gets postponed and sustenance projects use band-aids to treat bullet holes.
And that causes a trickle-up effect across the product companies selling into the industry - you don't sell new database licenses if billing systems aren't getting revamped and if that ripples continues, everyone upstream starts getting extremely frugal with their expenses.
Which is bad for the enterprise software ecosystem - the only folks who are likely to not notice that immediately will be small bootstrapped consumer companies, with real boxes to ship and lots of room to grow (or giant ones with room to grow, like a self driving car paid for by clicks).
I'm a bit apprehensive of drastic changes ever since I read the CA Budget (see last page of Economic Outlook[1]).
Oh yeah, I have my shortest stint in one of those company.
It was rumored that the company I was working had taken poison pill last time by having the ratio of benched workers and assigned workers to 50%:50%.
Their presence in our country was small, but they have multiple office campuses in India. One campus even have a separate building that if you walk to that building to work, people know you are benched.
Compared to other company that I worked, one of their value proposition is to have resources readily available if you want to start a project today, apart from having a dirt cheap offer. So yeah, benching a lot doesn't come as a surprise to me about those companies.
These companies hire a lot of graduates from all over the country at a merge rate and these newly recruited employees are then benched. They will sometime hire a couple of thousands of new grads from a particular school.
The bread and butter of many of those shops was already on the trailing edge when they came to the fore.
The market has dried up. There are springs of Windows, Oracle, NetSuite, Java and other Turn-of-the-Century technology, but new customers are fewer and fewer. Of those are a fraction who want/need bargain-grade Global IT Support.
Combine the declining presence of IT with manual administration interfaces, and it really is no great mystery.
>...newer areas such as digital and cloud, which require engineers to engage with clients instead of working remotely
Does that mean trends in remote work contribute to a measurable erosion of their collective markets? Or that their workforces cannot function in a face-to-face environment?
I am confused - Is this something new? As far as I can remember the perils of layoffs and afraid Indian IT professional is the "go to" story each year during the performance review cycle.
The only reason this is being played up is the perceived notion of protectionist environment throughout the world (not only US). Then there is the whole thing about "AI" and automation.
Or if you want to go full tinfoil hat on this - It is a ploy to stop people from asking better pay or searching for a new job.
Lately, I have been noticing the "increasing automation, use of artificial intelligence " line being touted by Indian offshoring companies whenever they talk about laying off or reduced recruitment. What exactly is the automation or artificial intelligence being used? I never hear any details.
Are they doing something radically new involving machine learning which allows them to be much more productive?
Or have they refined or streamlined their scripting environments and automated build\deploy tools? Perhaps they have built a suitably large and mature framework of components from previous projects which allows greater code re-use? In these cases, using words like "Automation" and "AI" is just putting a new sticker on an old product.
Or are companies just using these buzzwords to give a positive spin to an event whose real reasons are actually more mundane and financial(weak market conditions, reducing bloat in workforce ). Maybe they are attempting to mask something more ominous like significant shift in the market away from offshoring?
I suspect there's some truth to it, mixed in with the type of pulp fiction you identify.
For instance, consider the structural changes induced by the whole "cloud" phenomenon. In what is termed "legacy" infrastructure by the "cloud-minded", there is, classically, a whole caste (no pun intended) of operations and sysadmin people with a skill set of "installing servers", "keeping servers running", deploying software packages, applying patches, etc. A large stampede to cloud really does reduce a lot of this mundane but relatively skilled labour, in theory--it's one of the main selling points: lay off 80% of your sysadmin because the infrastructure mostly runs itself! To the extent that it doesn't, focus your operations efforts with a much smaller, leaner team of Cloud Rock Star Ninjas or whatever.
And "mundane but relatively skilled" is, as I understand it, the preferred area for classical Indian offshoring companies' labour arbitrage plays, since it's perceived to be the ideal target range, rightly or wrongly, for outsourcing by developed-world executives.
Then there's all the company that "cloud" has kept, whether conceptually or as a matter of timely coincidence, e.g. the slow fading of the distinction Corporate America makes between Operations and Engineering.
> What exactly is the automation or artificial intelligence being used?
Arguably continuous integration and deployment are forms of automation that eliminate developer jobs. It's not that there was a "run unit tests" job, but all developers got more productive so maybe a four person team could do the work of a five person team.
Or maybe that refactoring to migrate to the new architecture didn't require help from contractors anymore.
Nope, this is just a hype, as usual. The journalists are getting lazy, especially Indian journalists, they write anything like this for the sake of writing. There is no such thing as "due to automation and AI" revenues are down, on the contrary IF automation and AI is done, revenues and profits will go up! because of automation 4 people will be able to do work of 10 people. dumb journalism
> Are they doing something radically new involving machine learning which allows them to be much more productive?
Well they most likely not. But point here that they are losing lots of business so they would just not need that many people. If one read history it is the same pattern repeating first with textile, then component manufacturing in India. They think cost arbitrage is all it takes to become world leader. It works for some time and then labor saving techniques replace low cost labor.
I've seen a lot of these systems up-front. The reality is that most of them are RPA (Robotic Process Automation) tools. Vendors like BluePrism, Arago, and Automation Anywhere are some examples.
Longer term, they are trying to incorporate some learning algorithms to define a finite number of predetermined outcomes when those systems above spit out errors. A lot of this ultimately leverages error log analysis, known error databases, integrations to ServiceNow, and some scripts to take corrective action.
I have dealt quite a bit with these companies. They have great presentations about all the cool stuff they did for other companies and competencies they have. But when they get hired these competencies rarely materialize and you pretty much have a bunch of young engineers who don't know much. The senior architects who did the selling quickly disappear. I have never seen any of the advanced systems, processes and frameworks they are supposed to have actually being used at the customer site.
Traditional Windows system administration involves a ton of menial labor clicking through things according to procedures written down in Microsoft Word.
Maybe they are starting to operate more like tech companies with their Linux server fleets than SMEs with their Windows domains?
Dunno what the actual machine learning application would be.
I think this is the other way around: the world is moving towards machine learning and other complex stuff and Indian developers simply cannot keep up, given their average education and experience levels. Offshore development teams has always been of limited use for this exact reason, and now they are of even less use.
Exactly, I get irked when I read these stupid stories, there is no 'threat' of automation and AI, anyone who works in sw field knows the extent of manual work in it.
We are not even _near_ an age where we have working AIs sold to someone else. Watson is a real AI but it is an AI as a service, they don't sell watson clones and there is a reason, it is HUGE, hence you get the wit of Watson at a few $$$ per hour. This AI hype will go away soon and blockchain hype will start in a few months, "BLOCKCHAIN TO CAUSE MASSIVE LAYOFFS" "90% UNTRAINABLE IN BLOCKCHAIN"
these idiots don't stop for a moment and see if there is a need to train everyone in ML, if every dev in a company does ML who the f* will write other code?
Cognizant and Infosys are just the worst. We've been working with them on a number of deals lately where they are effectively middlemen in between us and a real customer who wants a product similar to what we provide. It's been a fucking shambles. Some days I'm just amazed at how much incompetence and incomprehension can be mustered on one team - everything is at least ten times more difficult than it ought to be, and more often than not they appear to be trying to deliberately sabotage the project.
Can someone give me some insight which sort of 'IT job' or 'programming' jobs are being replaced by automation? With a background in engineering, makes sense in manufacturing industry but in software development?
85 comments
[ 3.8 ms ] story [ 145 ms ] threadit also makes it easier to bring them later (after 1 year) into US on L1 without all these limitation of H1 like prevailing wage, yearly cap, spouse of L1 can work, GC seems to be easier (at least that seemed to be the case 10 years ago when I last paid attention to immigration)
Find some new, better friends.
Link to desktop site.
I'm sure you've heard about those people with 10 years of experience that are unable to code fizzbuzz? They hide out in those large corporate environments.
Specially taking into account the type of assemments being done to the employees, or which R&D unit is driving the main products (e.g. WinDev vs DevTools).
Yes, places like Microsoft System Research are great, but that is just a unit from the whole corporation. Same applies to the others.
After that they are perfectly fine not having IT people around.
The time each company had a set of developers to sit around and help them on as needed basis, is gone and won't be coming back.
Those that want to keep around onsite development need to upskill their social skills, and also be able to bridge between developer, architect and business relations.
I've never seen this work in practice. It always takes almost as many people to properly maintain a solution as it did to build it, there is no "done" state where the can get rid of the developers. It was accepted as far back as the mythical man month that maintenance is where the vast bulk of software costs are.
> Those that want to keep around onsite development need to upskill their social skills, and also be able to bridge between developer, architect and business relations.
At this rate we're about a decade a way from expecting developers to be one man corporations.
Those companies see software development as a cost and not something that need to be kept around.
Usually we also don't have someone sitting on our building to repair the doors when they get broken, we call someone to repair or replace them, when the need arises.
> At this rate we're about a decade a way from expecting developers to be one man corporations.
If you are living in an European country this has already happen long time ago.
In countries like Portugal, many companies are hiring developers on "Green receipts", which are literally one man corporations.
Similarly in countries like Germany and Switzerland, many consultants are freelancers, or work for a consulting company that plays the same role as agents do for sport and movie industries, giving them a fee. Also expected to be an one man corporation from the customer point of view.
The glorified idea of the developer that codes all day long on his cubicle/desk without anything else to worry about is long dead, at least in western part of the globe, when targeting the industries that don't sell software.
... accepted by companies run by people who read the Mythical Man Month.
Then there's the rest of the corporate world. These things you say are absolutely true, and yet not a day goes by that a company isn't learning that lesson, quite expensively--or better yet, failing to learn that lesson over and over.
The sad reality is that the way software is delivered at enterprise level, there is hardly any company that can be shown as example of excellence.
Only price and the output of applications matter, not how they are running inside.
Well the Indian IT companies have been growing for the last 30 years so they must be doing something right.
Many people managing software teams or budgets don't understand the coding business at all. Those people are like vegetarians running steakhouses.
I am not sure why are you so confused, that is how pretty much all for profit companies work or was it a poor attempt at sarcasm?
the parent points out that this market inefficiency exists because the people making purchasing decisions are not technically competent enough to truly evaluate the vendors that are selling them these development contracts.
A company can make money selling a bad product, it happens all the time; and its a well know phrase in finance that 'markets can remain irrational longer than you can remain solvent'
So you are telling me that these Indian outsourcing companies who have been growing for 30 years with billions of dollars in revenue is all because they are selling a bad product.
So the Indian companies are like wolves eating feeble Western companies like lambs. Do you realize how ridiculous that sounds.
Is it visible there?
It is such baseless stuff which irks me every time, I have read from MANY papers that automation is doing this that without explicitly saying X automation firm screwed up the revenues.
I would like to see visible proof that "automation" is eating into the revenues, no report says that, they just say "automation and AI"
The journalists don't know difference between AI vs ML and everyone is acting as if "AI and automation" will destroy X firm, no, they won't. Anyone who works in the sw field would know that there are many manual things done in the project so much so that there wouldn't be an "automation firm" which would sell blah blah robots to do human work. It is way off into the future, if it'll ever happen. This madness has to stop.
edit: clarification about who they meant
Yea, very true, The ratio of (Engineers / Managers) need to substantially increase over here.
I've seen it myself doing MSP work. Crappy old system breaks with every program update, every windows update, the wind blows the wrong direction, etc, etc. Client updates to new computers with SSD, new software based partially/mostly in the cloud. Full and correct integration with active directory. Their complaint load can easily drop 5 to 10 fold, mostly from less legacy software issues.
Each time something, which used to be done by humans, is done by a machine, that time spent doing that something is no longer payed to a human and only a fraction goes to the makers of the machine - that's the whole point of automation - reduce costs and increase productivity.
It doesn't lead to [more] humans being payed more money.
I think intuitively it makes sense, but coming up with undeniable proof is quite hard in the globalized economy - the effects of automation often happen further away in the supply-chain graph.
The 'modernized' company just stops importing or paying for X, because they can now do it 'in-house' for a fraction of what it cost them before.
The effect of that is that someone somewhere on the other side of the planet might not be able to find a job doing X as easily as they did before. Or their company goes bankrupt. Is it because of automation ? Could be.
It might create new opportunities, but those require newer and 'higher-level' skills - and as these opportunities arise, someone somewhere is already hacking away trying to automate them. And we're getting better and better at this, so the new opportunities are only available for a short period of time, until someone comes up with a brilliant new way of doing things. The software world is a great example.
I personally think it's very real and being in denial the way that many people in SV are, does not improve matters. We (the technologists) should at least ponder upon these questions and figure out solutions - otherwise we're leaving that in the hands of politicians, which serves as zero comfort to me.
A lot of these large IT companies relied on long term maintenance contracts for infrastructure and legacy software with unnecessarily large teams on self serving time and material (T&M) models. But these days, even large enterprises have to innovate to stay in the game and their experience with top end Agile consulting firms made them realize that you can do with much fewer people with automation and hiring better. On the other hand, the large IT companies got stuck with their old model and without training and experience on challenging projects, the talent atrophied even while having to eek higher salaries every year due to inflation and talent loss to startups and others.
However, the numbers being reported so far aren't huge compared to the employee sizes of these companies. For e.g. 5% layoff would actually be a reasonable trim on an annual basis. In that sense, the carnage hasn't started yet.
We need to separate our discussion points.
I know that tooling is improving considerably, and rightly so, but that doesn't mean that IT companies will now be downsizing to 10 because of chef or puppet.
If a company wants an inhouse dev team they'd do so without "automation and AI", the thing is, they don't and that's why enterprise software companies are formed.
Sure, I can bake my own bread, grow my own vegetables, mine coal for electricity myself, but I don't. I do what I do best and leave the other stuff to others. It is upto companies to decide if they want to build infra for them.
Whilst DevOps automation tools like Chef/Puppet/Ansible/Whatever have undoubtedly increased in popularity, very often they are just used as nicer ways to achieve the same thing IT professionals have been doing forever in shell scripts or similar.
This is purely speculation, but I wouldn't at all be surprised if US companies are simply less interested in hiring Indian outsource firms given the current climate and uncertainty around the future of work visas in the US. It's not hard to imagine such companies indicating they will spend less with these firms for the foreseeable future. I don't think anyone believes eventual reform of the H1B visa, the lifeblood of Indian IT firms in their biggest market, is going to be good for the Wipros of the world.
Exactly! The mgmt of the above stated firms sucks, they had no vision or a sense of direction, they make bad decisions and when they are asked questions like Sikka (infosys CEO was asked) Why do you earn millions per year and why do you take 50+% hike and still are unable to deliver then the hide behind "automation and AI" :)
it is mediocre CEOs and C*Os and a lot of other folks which are causing this downturn plus investors too, Google "eliott management letter" to get an indepth real reason behind this wave of "layoffs due to automation and AI".
You'd be surprised how many sys admins and especially developers have never written a shell script.
One of our code tests is to write a simple grep clone, the results seem to indicate that none have ever used a command line. Nearly every submission seemed to think that you could only pass one "-x" (x the placeholder) switch to a command.
More often than not, the bench exists to promise a customer that they have the people to "start yesterday" on the project.
The trouble with choking off the supply into the industry is that often the demand dies out as a result rather than increasing the price of a service - complete rewrites get shelved, new development gets postponed and sustenance projects use band-aids to treat bullet holes.
And that causes a trickle-up effect across the product companies selling into the industry - you don't sell new database licenses if billing systems aren't getting revamped and if that ripples continues, everyone upstream starts getting extremely frugal with their expenses.
Which is bad for the enterprise software ecosystem - the only folks who are likely to not notice that immediately will be small bootstrapped consumer companies, with real boxes to ship and lots of room to grow (or giant ones with room to grow, like a self driving car paid for by clicks).
I'm a bit apprehensive of drastic changes ever since I read the CA Budget (see last page of Economic Outlook[1]).
[1] - http://www.ebudget.ca.gov/2016-17/pdf/BudgetSummary/Economic...
It was rumored that the company I was working had taken poison pill last time by having the ratio of benched workers and assigned workers to 50%:50%.
Their presence in our country was small, but they have multiple office campuses in India. One campus even have a separate building that if you walk to that building to work, people know you are benched.
Compared to other company that I worked, one of their value proposition is to have resources readily available if you want to start a project today, apart from having a dirt cheap offer. So yeah, benching a lot doesn't come as a surprise to me about those companies.
The market has dried up. There are springs of Windows, Oracle, NetSuite, Java and other Turn-of-the-Century technology, but new customers are fewer and fewer. Of those are a fraction who want/need bargain-grade Global IT Support.
Combine the declining presence of IT with manual administration interfaces, and it really is no great mystery.
>...newer areas such as digital and cloud, which require engineers to engage with clients instead of working remotely
Does that mean trends in remote work contribute to a measurable erosion of their collective markets? Or that their workforces cannot function in a face-to-face environment?
Back in 2015 - https://www.wsws.org/en/articles/2015/02/27/indi-f27.html
2014 - http://indianexpress.com/article/business/business-others/la...
2013 - http://www.thehindubusinessline.com/companies/less-than-5000...
The only reason this is being played up is the perceived notion of protectionist environment throughout the world (not only US). Then there is the whole thing about "AI" and automation.
Or if you want to go full tinfoil hat on this - It is a ploy to stop people from asking better pay or searching for a new job.
Are they doing something radically new involving machine learning which allows them to be much more productive?
Or have they refined or streamlined their scripting environments and automated build\deploy tools? Perhaps they have built a suitably large and mature framework of components from previous projects which allows greater code re-use? In these cases, using words like "Automation" and "AI" is just putting a new sticker on an old product.
Or are companies just using these buzzwords to give a positive spin to an event whose real reasons are actually more mundane and financial(weak market conditions, reducing bloat in workforce ). Maybe they are attempting to mask something more ominous like significant shift in the market away from offshoring?
For instance, consider the structural changes induced by the whole "cloud" phenomenon. In what is termed "legacy" infrastructure by the "cloud-minded", there is, classically, a whole caste (no pun intended) of operations and sysadmin people with a skill set of "installing servers", "keeping servers running", deploying software packages, applying patches, etc. A large stampede to cloud really does reduce a lot of this mundane but relatively skilled labour, in theory--it's one of the main selling points: lay off 80% of your sysadmin because the infrastructure mostly runs itself! To the extent that it doesn't, focus your operations efforts with a much smaller, leaner team of Cloud Rock Star Ninjas or whatever.
And "mundane but relatively skilled" is, as I understand it, the preferred area for classical Indian offshoring companies' labour arbitrage plays, since it's perceived to be the ideal target range, rightly or wrongly, for outsourcing by developed-world executives.
Then there's all the company that "cloud" has kept, whether conceptually or as a matter of timely coincidence, e.g. the slow fading of the distinction Corporate America makes between Operations and Engineering.
Arguably continuous integration and deployment are forms of automation that eliminate developer jobs. It's not that there was a "run unit tests" job, but all developers got more productive so maybe a four person team could do the work of a five person team.
Or maybe that refactoring to migrate to the new architecture didn't require help from contractors anymore.
Well they most likely not. But point here that they are losing lots of business so they would just not need that many people. If one read history it is the same pattern repeating first with textile, then component manufacturing in India. They think cost arbitrage is all it takes to become world leader. It works for some time and then labor saving techniques replace low cost labor.
Longer term, they are trying to incorporate some learning algorithms to define a finite number of predetermined outcomes when those systems above spit out errors. A lot of this ultimately leverages error log analysis, known error databases, integrations to ServiceNow, and some scripts to take corrective action.
Traditional Windows system administration involves a ton of menial labor clicking through things according to procedures written down in Microsoft Word.
Maybe they are starting to operate more like tech companies with their Linux server fleets than SMEs with their Windows domains?
Dunno what the actual machine learning application would be.
We are not even _near_ an age where we have working AIs sold to someone else. Watson is a real AI but it is an AI as a service, they don't sell watson clones and there is a reason, it is HUGE, hence you get the wit of Watson at a few $$$ per hour. This AI hype will go away soon and blockchain hype will start in a few months, "BLOCKCHAIN TO CAUSE MASSIVE LAYOFFS" "90% UNTRAINABLE IN BLOCKCHAIN"
these idiots don't stop for a moment and see if there is a need to train everyone in ML, if every dev in a company does ML who the f* will write other code?
XRP (ripple) seems to have the potential to disrupt a few thousands job in banking in the near future
Eastern Europe seems to be quite popular.
Avoid terrible resellers and body-shops.