‘GenAI programmes may not be large in terms of value, but have triggered a lot of new opportunities among clients.’
Kindly note the image has only been published for representational purposes. Photograph: Piroschka van de Wouw/Reuters
India’s third largest IT services player HCLTech’s Q2 FY25 numbers beat estimates on growth and profits.
CEO and MD C Vijayakumar sounded confident about the macro situation, saying discretionary spend was better in Q2 than Q1.
In a video interview with Shivani Shinde/Business Standard, Vijayakumar talks about growth triggers, spend mindset and GenAI deal growth.
You have revised the revenue guidance up on the lower end. What stops you from upping the higher end of growth guidance?
We have to be cognisant of some macro and geo-political issues and also some client-specific issues like automotive and aeronautics.
So, those are real issues and we have to keep that in mind. I think the market still has a mix of cost optimisation and some change focused programmes, and a lot more interest in GenAI and the art of possible.
Photograph: Kind courtesy HCLTech
You also said discretionary spend in Q2 was better than Q1. Also, the US Federal Reserve rate cut also came in. What are your views?
Yes, discretionary spend is picking up. I do not know if there is a direct correlation with the Fed rate cut, but obviously we saw good traction in a number of verticals in the last (Q2) quarter. And, some of that had started even towards the end of Q1 FY25.
There are, of course, efficiency-led programmes, but even in these deals, many are trying to use new capabilities like GenAI.
For instance, how does one deploy it, scale it and where can it make a difference to run programmes. These are the areas where discretionary spends are happening.
IMAGE: The HCL Technologies office in Noida. Photograph: Mansi Thapliyal/Reuters
How are GenAI-led deals panning out? Are they out of the proof of concept (PoC) stage and now into scale?
GenAI programmes may not be large in terms of value, but have triggered a lot of new opportunities among clients.
For instance, if someone has a lot of technical debt, which most of the Fortune 500 firms have, they want to find the right opportunity to modernise the landscape. And, GenAI is very helpful in that process.
Similarly, once you want to scale, getting the data strategy right becomes a very important element.
The third, which is still early, is the cognitive infrastructure, where customers would look at hyperscalers and private AI stacks to deploy GenAI solutions. We are seeing scale in some pockets.
How is HCLTech’s AI platform being adopted among clients?
Our goal is to embed the AI Force platform for service transformation across all our customers.
We have started with a small list of 50 clients and almost half of them are now using AI Force.
AI Foundry helps in streamlining the data and creating the right data pipeline. And in GenAI Labs, we bring customers with an idea or do a PoC and create a minimum viable product for clients. These are the three motions and all of these bring opportunities.
IMAGE: C Vijayakumar.
Photograph: Kind courtesy HCLTech
Q2 also saw the software business scale. Is it big enough now to impact the overall performance of the company? What is driving this?
It is still 10 per cent of our revenue. Right now, this business is growing in the low-single digit.
We want to get it to mid-single digit. That is our near-term aspiration.
A lot of product features that we have developed over the last few years are making them more modern.
So, customers can adopt this with any hyperscaler of their choice. Or, can adopt them in a private Cloud environment too.
The product capabilities are focused on existing customers and make deployment successful. More importantly, showing empathy to customers also helps.
Your HR head had said that even fresher hiring will be focused on specialisation. How will this impact the pyramid structure?
The skill demand units are changing and when you hire people with specialised skills, the cost structure will be different. Hence, the pyramid model will also change a little bit over a period of time.
Feature Presentation: Rajesh Alva/Rediff.com
