As client conversations increasingly revolve around artificial intelligence (AI) and enterprises move from experimentation to deployment, Tata Consultancy Services (TCS) expects demand to improve despite continued weakness in consumer-facing sectors. TCS Managing Director and CEO K Krithivasan speaks to Urvi Malvania about the demand outlook, AI-led transformation deals and the growing shift towards outcome-based delivery. Excerpts:
What gives TCS the confidence that the September quarter will be better than the June quarter?
Customer discussions across industries, the total contract value (TCV) we signed and feedback from our client partners give us confidence. We expect manufacturing and life sciences to recover after a weak first quarter, while banking, financial services and technology services should continue to perform well.
Do you expect stress in the consumer vertical to continue? Is geopolitical uncertainty in West Asia still affecting demand?
The consumer business is closely linked to geopolitics. Airlines are under pressure and, because of inflation, discretionary retail spending remains weak across several geographies. As geopolitical conditions stabilise, spending and investments should resume. The direct impact of the West Asia crisis on us is limited, but secondary effects such as pressure on European airlines, tourism and fuel costs are influencing demand.
You said there has been a marginal increase in AI transformation deals. How is the deal pipeline evolving?
We signed more net new deals this quarter than in the previous one, which is encouraging. Today, almost every customer conversation and every deal has an AI component, either to improve productivity or create business value. That AI component will continue to grow over time.
The chairman has spoken about having as many AI agents as employees within three years. How do you plan to achieve that?
We already have around 2,500 AI agents delivering customer solutions. Employees have also built many more for their own productivity, perhaps 10 times that number. Going forward, there will no longer be a linear relationship between the volume of work we deliver and the number of people we employ. AI agents will augment employees, enabling us to deliver significantly more work with the same workforce.
AI models come with high token costs. How are these costs reflected in deal pricing?
The productivity gains we deliver today already account for token costs. Customers ultimately look at faster delivery, better quality, lower total cost of ownership and a positive return on investment. There is a perception that every workload requires the latest frontier model, which is expensive. In reality, workloads can run on open-source models, on-premise deployments or earlier-generation frontier models.
Selecting the right model for the right workload is where a system integrator creates value, supported by FinOps capabilities that optimise costs. Commercial structures vary. In some engagements, token costs are bundled into the pricing, while in others, customers pay for tokens separately.
What is the latest update on TCS’ planned data centre investments?
The project is progressing well and we are evaluating multiple locations. We are in advanced discussions and expect to announce a definitive Letter of Intent during the second quarter. The infrastructure will be spread across multiple states rather than a single location, and the announcement will be made along with customers that have committed to capacity.
How much of TCS’ business is now outcome-based?
A. It depends on the type of engagement. Many Global Business Services deals are already outcome-based. In IT services, the commercial model may not explicitly be outcome-based, but we often commit to delivering specific productivity improvements, such as a 15% gain, and execute against those commitments. So, even if the billing model is different, a significant share of our engagements today is built around delivering committed business outcomes.