Scale isn't sexy anymore: How mid-cap IT challengers are winning the AI era

New Delhi, July 29 -- Persistent Systems, an Indian technology services company, had an eye on the German firm Nagarro for a few years.

The digital engineering firm with over 300 customers in Europe first came into the limelight after private equity player Warburg Pincus explored a potential buyout in late 2024-the Munich-based company's board rejected the offer.

That footprint in the continent also interested Sandeep Kalra, the executive director and chief executive of Persistent Systems. In June this year, the German company's board supported a transaction proposed by the Indian company and an agreement was signed. Nagarro would combine with Persistent Systems for $1.3 billion.

Kalra's company secured about a 21% stake in Nagarro through an agreement with its founding shareholder and launched a voluntary public takeover offer for all remaining shareholders. If the acquisition goes through, it would add over a billion dollars to Persistent Systems' revenue-the company generated $1.7 billion in 2025-26.

"We have been competing with larger players. We had the capability, the agility, but lacked the global footprint," Kalra told Mint over an MS Teams call. "Now we have the nimbleness of a smaller company and the turf has equalized on the scale and presence sides."

A few months earlier, Coforge made an even bigger bet, acquiring California-based Encora in a $2.35 billion deal, which the company claims has enabled it to "create a $2 billion artificial intelligence (AI)-led engineering, data engineering, and cloud engineering platform." Mphasis, meanwhile, acquired Theory and Practice, an AI decisioning company from Vancouver.

In regular times, these deals might have been seen as business as usual. Coming as they are in a turbulent, AI-led disruptive era, they are more than just any buyout. They point to a newfound confidence among mid-tier IT services providers to scale their operations.

Mid-tier companies are also growing faster and seem willing to take more risks than their larger peers.

In 2025-26, the large firms-TCS, Infosys, Wipro, HCLTech-together generated around $80 billion, but growth was muted, ranging from flat to mid-single digits. In contrast, mid-tier players-Coforge, Persistent Systems, Mphasis and Hexaware-delivered around $7 billion in revenue, growing at double-digit rates, with Coforge surging nearly 30% compared to the previous year.

By pivoting quickly to AI-led delivery and outcome-based contracts, these firms are winning $50-100 million deals once reserved for the giants. And right now, this cohort appears better placed to ride the disruptive AI wave.

For years, India's information technology industry was defined by scale. Bigger meant better. The largest firms won billion-dollar outsourcing contracts by offering large teams of engineers, global delivery centres and the ability to execute sprawling, multi-year projects. AI is rewriting that playbook.

"The traditional linkage between headcount and revenue is already breaking down," said Nitin Rakesh, chief executive at Mphasis. "In an AI-led environment, we are not solving for how many people we deploy, but for what outcomes we deliver."

According to market sources, in recent years, mid-market companies have won business from Citigroup, Cigna Healthcare, Sabre Corp, Charles Schwab, Allianz, Nestle and Ryanair. In fact, Coforge bagged its largest-ever mega contract, a 13-year deal worth $1.56 billion from Sabre Corp, a travel technology company, competing against top-tier providers.

Global customers are expecting more AI agents to be deployed and fewer people. Software development itself is changing, with AI automating code writing, documentation and testing while human engineers focus on architecture, design and business context.

For example, consulting firm Accenture built a 'gen AI factory' for Air France-KLM in 2025 to automate customer support and draft e-mail replies, freeing up human employees to do higher-level operational supervision like fraud investigation or flight disruption management.

That shift-from selling effort to selling outcomes-is playing directly into the strengths of mid-tier IT companies. And they are responding by changing the people pyramid.

Rakesh says the model naturally shifts to a more balanced, non-linear 'diamond' shape, with greater emphasis on capability than sheer scale.

Diamond shape refers to the change in hiring trends that have affected the shape of the workforce pyramid. Junior-level engineering roles, typically referred to as L0 and L1 (for instance help desk support engineers), have become 'human-less' as AI agents have taken over. The L2 layer (like advanced technical support) is assisted by agents.

One peculiarity of the AI era is a change in client buying behaviour.

Historically, Indian IT firms competed for massive outsourcing contracts running five to seven years. AI projects increasingly look very different. TCS recently acknowledged that AI engagements are shorter. This requires the company to continuously win new engagements to sustain growth.

"AI revenue is not like traditional ADM (application development maintenance) revenue where there is a lot of annuity revenue involved. Many of these projects tend to be one quarter, two quarter projects, where we complete and we have to win new projects again to deliver and accrue the revenue," K. Krithivasan, CEO of TCS, said during the company's first quarter earnings call in July this year.

This trend plays into a market where companies like Persistent, Coforge, Hexaware and Mphasis have traditionally operated. These companies have always thrived on shorter-cycle deals.

"They (mid-tier IT) are surely adapting to AI disruption and opportunities better than the larger peers. This involves strengthening their platform story, investing in select niche talent and scaling their key strategic partnerships," said Yugal Joshi, partner at Everest Group, a research firm.

Unlike traditional IT services, where every project starts from scratch, AI platforms allow companies to reuse intellectual property across multiple clients. Every new implementation makes the next deployment faster and potentially more profitable.

Rather than pitching AI as another software tool, mid-tier firms are embedding it into reusable platforms that can be deployed repeatedly across clients. Mphasis has developed NeoZeta, its AI-powered platform to modernize legacy applications, along with DeepInsights, which uses AI to help enterprises extract business intelligence from data. Coforge has built AgentSphere, a platform for deploying autonomous AI agents, while also rolling out more than 120 'ModSquads', which combine human engineers with AI agents to execute projects. Persistent has been investing in its SASVA platform to accelerate enterprise AI adoption.

To be sure, large IT firms also have AI-focused platforms. Some examples include AI WisdomNext from TCS, Topaz from Infosys, Wipro's ai360 and HCLTech's AI Force.

The first wave of generative AI projects focused largely on experimentation-chatbots, copilots and coding assistants. Enterprises are now entering the second phase, where AI is expected to automate business processes, redesign workflows and improve decision-making. This requires companies that understand both technology and industry domains.

Here the mid-tier providers' specialization in a few verticals gives them an advantage.

Many smaller firms spent the last decade building modern digital systems rather than maintaining legacy ones. Persistent, for instance, built expertise in product engineering, healthcare and banking. Coforge developed strong franchises in travel, insurance and transportation. Mphasis sharpened its focus on financial services.

"The future belongs to firms that combine AI with domain expertise," Ramkumar Ramamoorthy, partner at Catalincs, a tech advisory firm, said. "Technology alone is no longer enough," he added.

Meanwhile, mid-tier IT firms can move faster because of their leaner structure. Investor expectations are also different when compared to large companies.

"Large firms simply carry too many service lines, layers of leadership and a heavy load of legacy systems," Gaurav Vasu, CEO, UnearthInsight, a consultancy, said.

The top four IT services companies have historically delivered 15-18% operating margins and the Street expects them to sustain this. Mid-tier firms have historically operated in a 12-15% margin range. And so, the quantum of failure risk and investor pressure is different for both.

"Mid-tier firms' boards and investors tend to be more risk-taking while large firms lean on conservative measures-favouring buybacks and dividends," Vasu said.

While India's large IT companies continue to trade largely on stability and cash generation, several mid-tier firms have increasingly been valued for growth potential and a faster pivot to AI. The market is effectively betting that smaller companies can reinvent themselves quickly.

Brokerage firm Elara Securities expects Persistent Systems and Mphasis to lead in revenue growth for IT services in the coming quarters. Kotak Securities, meanwhile, noted that "challengers will continue to outperform incumbents". TCS among incumbents and Coforge and Hexaware among challengers are its preferred picks.

Over the past year, Indian IT stocks experienced a widespread downturn. Industry giants bore the heaviest losses, with Infosys down 32.58%, Wipro down 32.47%, and TCS dropping 28.60%, HCLTech at negative 16.12%. Mid-cap software players managed to marginally outperform their tier-1 peers (and the Nifty IT index which fell around 20% in the year), but still finished the year in the red; Mphasis declined by 14.97%, Coforge dropped 14.11%, and Persistent Systems remained nearly flat with a minor loss of 1.03%.

Though the current market situation favours asset-light mid-tiers, it's not without challenges. The same characteristics that make them agile also expose them to greater risks. Many remain far more dependent on a handful of large clients than the industry's biggest players. A delayed project or a cancelled transformation programme can have a disproportionate impact on quarterly earnings.

Joshi points to their poor profitability as an Achilles' heel. "Most have over 500 bps difference compared to their larger services providers," he said.

Even as deal sizes are changing, competition is also changing business models. TCS is building a team of 8,900 forward-deployed engineers (FDEs) and is scouting for new AI deals. An FDE is a software engineer who works directly inside a client's environment to customize, integrate, and troubleshoot technology. Unlike traditional engineers who work entirely at their own company's headquarters building products, FDEs get 'deployed' on the front lines to ensure a product works in the real world.

In its recent quarter earnings in June, Accenture said it is going to focus heavily on the mid-market segment-companies with revenue between $300 million and $3 billion-through its new platform called Accenture Edge. This is also the market where mid-tiers have had an advantage and now they face competition from a global giant.

There's also growing competition from AI native providers like Cognition Labs Devin, Anthropic Claude Code and GitHub Copilot. These are like autonomous software engineers which can run whole IT projects.

For three decades, India's software story was built on one simple equation: more engineers meant more revenue. The small players around Y2K-TCS, Infosys, Wipro, Cognizant, HCLTech-became big, leaving a long tail of also-rans. Some of them have scaled now to form a club of mid-tiers.

"Every disruption brings its own 'S-curve' of newer players and challenges for incumbents," said Joshi.

The S-curve here represents the lifecycle of new technology. Big changes (like AI) start slow, accelerate fast, and usually allow newer companies/start-ups to replace giant, old-school companies that refuse to adapt.

"More than two and a half decades back, when the offshoring wave started, not every IT services firm made the transition, and they continue to struggle to this day. Similarly, the AI wave will create its own winners and losers," Joshi concluded. Published by HT Digital Content Services with permission from MINT. For any query with respect to this article or any other content requirement, please contact Editor at [email protected]