AI is coming for India’s code factories

It’s not easy to avoid talking about artificial intelligence or AI nowadays if you are interested in what happens to your portfolio. Many chipmakers that manufacture chips that go into running many AI-based models or services like ChatGPT require enormous computing power. Right now companies that make these chips like NVIDIA, Advanced Micro Devices (or AMD), Intel, among others, are in focus because without these semiconductor chips, there can be no AI revolution.

Over the past few weeks, the global stock markets seem to have fallen out of love with these companies as many AI chip makers’ stock took a beating. So does this mean the dawn of the AI revolution is over?

Not if you are watching or reading what the top bosses at many Indian IT companies are telling their shareholders, and analysts who track these companies, during conference calls. The mention of AI was customary in these calls till a few quarters ago. But as the popularity of AI-chatbot ChatGPT rose over the past year and a half, the use of the term AI in these calls also increased dramatically.

But is the $250 billion Indian IT services sector ready to adapt to this new reality? Right now, companies like TCS, Infosys, Wipro, etc, are talking about small pilot projects that they are working on with their clients using AI. They are also talking about training their staff in ‘AI technologies’. It’s essentially an in-house training course that tries to impart skills to staff for future projects.

Will the Indian IT sector be able to adapt in the face of this revolution? Will they struggle to offer new-age services to their clients??

To understand what is happening now, we will have to get into a short history lesson about the evolution of the Indian IT services industry. Let’s get into it.

Evolution of Indian IT industry

India’s IT services industry came into its own more than two decades ago when the world was grappling with the Y2K bug. This was back in the 1990s when the internet and computers were being adopted by people all over the world. The Y2K bug was expected to cause havoc in computer storage and data formatting as computer systems used to store the dates as ddmmyy instead of the post Y2K syntax of DDMMYYYY.

Indian software engineers made a name for themselves by ensuring global companies could deal with the Y2K or millennium bug. They wrote lines of code to ensure systems could deal with the 4-digit year data storage for smooth transition into the new millennium.

Over the 2000s, India’s IT services companies expanded their services across the world, into various industries, verticals and sectors. Indian IT service exporters lapped up outsourcing deals from many businesses across the world. Those businesses who didn’t want to run large IT operations in-house chose companies like Cognizant, TCS, Infosys, Accenture, Capgemini, among others.

The first decade of the new millennium saw engineers being hired and trained how to code (the code factor), and then sitting on the bench, waiting to be deployed on projects. Each offshore engineer being deployed in projects was a billable asset. That is what led to large training centres being set up by companies like Infosys across India to train people to code.

These engineers also used to be deployed onsite at the client’s premises across the world. The costs were higher to deploy an engineer onsite because of higher travel and visa related costs. But many did so because staff had to be present on premises to manage the systems and any integration issues.

One context that will help understand this paradigm is the existence of large physical servers to run organisation-wide systems to manage various operations in manufacturing, BFSI, pharmaceuticals, oil & gas, and many other industries.

Also during the 2000s, many companies moved into consulting projects for implementing and managing large IT systems during the 2000s to diversify their businesses. After the Global Financial Crisis in 2008, the Indian IT sector was in doldrums as clients held onto their purse strings because of an uncertain economic outlook. It took a while before things started to look up, but then came another upheaval.

The digital revolution

The 2010s saw the rise of digital technology. Cloud-based servers meant that the traditional physical servers-based systems needed to be upgraded. Indian companies didn’t have much experience in this. Some analysts and brokerages downgraded many Indian IT services companies to a sell rating because they weren’t sure if they had the necessary knowhow to offer these services to their clients.

Another thing that came with the digital revolution was the reduction in contract durations. The older era saw large multi-year deals being signed, and this changed dramatically. Even if the deal size was large, say above $100 million, they weren’t for 7-8 years, but for just a couple of years, or sometimes a little more.

The first murmurs of AI and automation changing the computer engineer’s job started in the middle of the 2010s. The cloud-based servers distributed computing power remotely and companies just needed to find better ways to use this power to run various business processes. This distributed computing power would eventually lead to the likes of ChatGPT and other similar large language models in the 2020s.

Essentially, there was an inevitability to the AI age. As cloud computing allowed businesses to rent computing power based on their needs, cloud services providers like Amazon Web Services, Microsoft’s Azure, Google Cloud, VMWare, etc started to serve this need.

The AI age beckons

You know you are in a paradigm shift when the headlines start predicting doom when it comes to people and their jobs. Just do an internet search on AI and jobs and you will find yourself scrolling furiously through a bottomless webpage. There are so many predictions that it is difficult to understand how this could affect Indian IT services companies.

Companies like TCS, Infosys, Wipro, and many others are talking about how they are trying to work with clients to increase AI use cases that they can offer along with their existing products and services. Infosys called out during its Q4FY24 post earnings conference call that it experimented with a large language model generative AI tool that wrote 3 million lines of code. As of now, the company says it managed to retain this automated process in-house. Whether this will continue to happen in the future for the entire industry is anybody’s guess.

The age-old question of whether Indian IT service companies will graduate towards the cutting edge of technology is more relevant today than ever before. The industry hired ahead of demand during the pandemic, and now is seeing their headcounts fall year-on-year.?

Their clients are not taking up non-critical projects, and only focusing on short-term benefits while spending money to upgrade their technology stack. Many of these clients are also looking at acquiring smaller AI startups to experiment with use cases themselves.

Cloudy skies

As the earnings of more IT companies come out, the revenue outlook looks grim. For years, they have undercut global IT services providers like Accenture and Capgemini to bid for projects because of the cost advantage the lower salaries for Indian engineers provided. Employee costs make up 50%-60% of most IT companies’ revenues, but still allow many of them to earn net profit margins ranging from 10%-20%.

In an era where engineers’ salaries are elevated, it would be foolhardy for Indian IT companies to not dabble with AI to deliver their services. At the same time, they are also training their existing staff in technologies that will help them harness the power of generative AI large language models to deliver their services to clients.

That is what is happening right now. Clients are asking for it and these companies have to figure out a way to deliver. The AI era has come in after 2 consecutive years of tepid revenue growth (ending FY24) is compounding the problem. No wonder that the consensus view on Dalal Street on many IT services companies is either to Hold or Sell these stocks. (Do consult your financial advisor before taking any investment decisions.)

The point to all of this is to illustrate that there have been many epitaphs written about Indian IT, but the industry has learned to pick itself up after being knocked down. The question is whether it would be able to do it again in the AI age in double quick time.

Either way, the code factory era of Indian IT services is seeing its last days in the sun.