When do you see this low single-digit growth environment changing? Hiring too has been very weak. Is it primarily because of global macro conditions?
I think employee growth is going to be very strong. GenAI is giving us new areas where we can do projects. So, while there are some areas for improvements, overall, we see a positive revenue impact on growth. We’ve called out a guidance of 3% to 4% for the full financial year. We don’t have a view at least of what the next financial year will look like in terms of growth. Broadly, we have extremely talented people in India, so I don’t see how anyone can really compete with that.
The American economy is seen to have been relatively strong for some time. Yet, there is this demand slowdown for IT…
Large tech projects need big commitments of capital as you start those programmes, and given the interest rate increases in the last two or three years in Europe and the US, we’ve not seen such large programmes starting. My sense is companies are being careful. Right now, a lot of the work we see is around cost takeout or efficiency.
In 2015, Infosys had funded OpenAI and ANSR (GCC consultancy and solutions company). But it subsequently did not take that forward or divested. Today, the two words everyone in the Indian tech industry are talking about are AI and GCC. Very recently, Accenture invested a big amount in ANSR. Did you miss the bus on those two opportunities?
Those, as you know, happened a while ago. If you look at generative AI, what we are doing today is market leading, and we are quite happy with the outcome there. Regarding GCCs, there are multiple themes. First, we are partners with them as they are growing in India, because many of their parent organisations have been clients of ours for years. Second, we’ve seen that these GCC investments have a lifecycle. Things that were created 10-15 years ago, with changes within the companies, they want to move to some different areas, and then they move away from it. So, in the last 2-3 years, we have taken over some GCCs which are older. At the start, we help and support, and as they mature, because of our scale, we are much more able to bring efficiency to those programmes. The automation we build is very different from what some of the GCCs employ. So, as they grow in the maturity lifecycle, we see clients saying, can you take that over and make it more efficient, or become partners with us.
There are GCCs that say it’s more efficient to do tech in a GCC, than with an IT services partner. Several seem to be bringing more and more work in-house. They also seem to want to do cutting-edge tech like GenAI themselves, instead of giving it to IT services players. They seem to be taking over the higher value work.
When a new technology comes, there’s a lot of interest with the client to build something which is more within their own organisation. What is the benefit with us? We are doing hundreds of projects in GenAI, across every industry. Our experience, the people that we work with, the knowledge that we build will be much broader based. So, it’s very difficult for someone else to have that level of expertise and depth. Because it’s new, it’s something that many large companies want to make sure that they have some ability to drive themselves. But as you build scale, that knowledge and depth is where we are very good. And we’ve seen over time that this plays out. As for efficiency, newer GCCs are probably one hundredth, if not one-thousandth of our size. It is difficult with that scale to make the sort of efficiency play we have, or even the automation tools that we use.
How do you see your progress with GenAI? Where is it having the biggest impact?
In GenAI, we are seeing a real expansion of work. We have 270,000 people trained on GenAI, what we call GenAI aware, practitioners and experts. We are working with something like 16 different large language models. There are three big areas where we see impact. One is software development, the second is customer service, especially in operations and tech. And the third is knowledge objects – we’re working with a bank in the credit area, with a pharma company in their drug discovery area, and with a telco in their product specification area – so anywhere there’s knowledge objects where you can create some value from faster or more insightful ways of looking at that knowledge. We’ve been using these things ourselves to improve what’s going on within Finacle (Infosys’s banking product). And there are several tools we built. One example is, we’ve taken an open-source large language model, then built a tool for ourselves on it where we can do more efficient work with Java development on a desktop. So, it completely changes the game on what we can do. We rethought all our service lines with generative AI.
Can you provide metrics of improvements in areas where GenAI is being applied?
In software, we are seeing in the range of 15% benefits in what we are developing with our clients. In customer service, we are seeing great benefits in HR functions.
Nilekani spoke about how AI doomerism has quietened down. With his tech evangelism and boardroom connections and your soft power leadership, how are you rallying forces, especially on the sales and delivery side, to become an AI-first enterprise?
Nandan is a tech visionary and he’s obviously very active with us. We are fortunate in that. We want to be more aligned with where clients are going, and that was the change we made. AI-first is understanding that clients want to make changes faster internally. Internally, we have an app, InfyMe, which all of us use for internal things. We have an AI assistant that helps us, whether it’s with sales, delivery, HR, or training. We feel quite good about this. We want to make sure we’re doing what clients are looking for. A few years ago, there was a much stronger move to digital and cloud. That is still very much part of it. Today, there’s much more of an interest in GenAI, and that’s how we’re changing our company. Our view is that technology will change in five years and Quantum might be the big thing. We will be more aligned to that because that’s what our clients are looking for.
Infosys is in the leaderboard quadrant in terms of use cases for GenAI. Some of your peers have called out the pipeline for GenAI, even though it’s not converted into revenue yet. When do you think you’ll move away from POCs to seeing revenues trickle into the books?
The revenues are already beyond the PoCs. They’re not just trial pilots. They’re not large projects, but they’re real projects with real impact. We have already been tracking what is generative AI, revenue and so on. At some stage over the next few quarters, we will start to call it out externally. I think a lot of things can be broadly put into AI. So, we are careful to make sure we are measuring GenAI the same way we did with digital. Every project is marked internally, so that we know it’s an auditable process.
But the split between traditional AI and Gen AI, do you think there’s value in calling them out separately?
Clients view GenAI as transformative. There are more revenue growth opportunities because you can go faster into some markets. You can do cost efficiency because you can do things better. So, there are three big advantages – revenue, cost and time. When we started digital, the cloud was one piece of it. But today it is the biggest piece in digital.
How comfortable are enterprises with chatbots? There are hallucination issues.
The cleaner the data that is training it, the lesser the hallucination. But in normal life, we hallucinate – forget GenAI, we as people, when we are interacting, we may make some mistakes. So, there is nothing unusual about that. The idea is to reduce and not let that error contaminate other things. Today, there is no usage which is coming to what is called critical system usage, because people are waiting to see how the hallucination can be reduced further. But if you’re doing it in different areas — error mapping and making people more productive — then you’re already in a better place.
How do you position yourself against the big four? They straddle the entire spectrum, while companies like yours predominantly do technology work.
We don’t see them when we compete in the tech world. I think our strength, as you said, is we are very tech oriented. So, if you’re a CIO looking to make a choice on who to give your technology work to, I think there are very few options outside of Infosys. Very few CIOs are going to pick someone who is not as strong as Infosys in the tech area. For large companies looking to get technology delivery done with the lowest risk, I think Infosys is the best choice going forward.
Cybersecurity is a boardroom agenda. In your annual report, you’ve disclosed how Infosys is under-covered for cybersecurity insurance. Do you want to increase that coverage?
In cybersecurity, we believe we are well covered, but we must highlight all the risks cybersecurity inherently has. We have security operations centres in multiple locations around the world where we support clients. Cyber is a growing area where we see clients have more needs. But there are also continuous threat actors which we must be watchful about.
Is there a truce after the GST pre-showcause GST tax notice?
We gave an update a few days ago with the disclosure in the market. There is no new update from that time.