The head of Mphasis believes that discretionary tech spending is rising
Tech Spending to Accelerate in 2026 as Enterprises Scale AI, Says Mphasis CEO
Discretionary spending on technology is expected to accelerate in 2026, even as enterprises aggressively reprioritise budgets to fund artificial intelligence (AI) initiatives, according to Nitin Rakesh, chief executive of mid-tier IT services firm Mphasis.
“As enterprises move past macroeconomic concerns and better understand how to embed AI into their businesses, investment in technology is only going to increase,” Rakesh told ET in a virtual interview. He noted that while tools like ChatGPT sparked excitement when they emerged in 2022, enterprise adoption of AI takes time and careful execution.
Rakesh said AI is becoming deeply embedded in enterprise technology stacks, forcing companies to rethink how limited capital is allocated. “We have validated proof-of-concepts and use cases. The focus is now on scaling—deciding which data models to use, what AI stack is most suitable, and how data governance will work,” he said.
According to Rakesh, enterprises are identifying specific areas where AI can drive efficiency by automating tasks that require limited manual intervention, including parts of business operations and software engineering. Large transformation deals increasingly centre on data centre consolidation, cloud migration, application modernisation, and decisions around GPU versus CPU usage for AI workloads.
“It’s not that clients want fewer engineers,” he said. “They can simply write more code. There is always more coding to be done in the backlog.”
The rise of AI is also reshaping commercial models in the IT services industry. Rakesh said traditional labour-arbitrage models—based on headcount and hourly billing—are giving way to outcome-based pricing structures where AI agents work alongside human teams. Mphasis’ fixed-price contracts have grown nearly 50% year-on-year, as clients re-bid deals early to capture AI-driven efficiencies.
On the technology front, Mphasis has chosen to build its own agent layer while partnering with providers of foundational large language models. The company uses Google’s Gemini as its native model and also works with Anthropic and OpenAI. “We don’t see why we need to reinvent that wheel,” Rakesh said, adding that the firm’s intellectual property lies in training and embedding AI agents into delivery workflows.
This platform-centric approach is also guiding Mphasis’ acquisition strategy, with a preference for buying IP-rich platforms rather than pure services companies.
The shift to AI is changing workforce strategies as well. Mphasis plans to hire about 500 students from a hackathon conducted with Plaksha University as “AI natives,” engaging them up to a year before joining to ensure rapid deployment on AI-intensive projects. Internally, the company runs a platform called Gig Cloud, allowing employees to take on additional work beyond regular client shifts. About 80% of staff are on the platform, with roughly 20% actively participating, Rakesh said.
While some roles—such as testers and production support engineers focused primarily on monitoring systems—may decline, demand is expected to grow for professionals who can build, troubleshoot, and orchestrate complex systems. “The reshaping of the IT workforce is about separating functional roles from deep technical skills,” Rakesh said.