India’s GCCs shift from delivery arms to AI buying centres, requiring partners to reset strategy

As R&D, engineering, and AI CoEs gain independent budgets in India, partners must realign for a buyer-led shift inside GCCs.

GCCs in India are no longer just execution engines for global headquarters. They are increasingly turning into budget owners for AI work, with independent mandates spanning R&D, engineering, and AI centres of excellence. That shift is changing the buyer map for the entire services and partner ecosystem, and forcing a rewrite of how firms build coverage, capabilities and account strategy.

Speaking to CRN India, Rakesh Sancheti, chief growth officer at Tredence, positioned this evolution as a structural change in how global organisations operate in India, not a short-term cycle driven by cost arbitrage. He said the language itself is changing - from GCCs to what many now call “global value centres” - reflecting a shift from delivery to value creation.

If decision-making and budgets move closer to India-based GCC leaders, partners that still treat India as a downstream delivery layer will miss where the next wave of AI spending is being approved.

From execution arm to buying centre

Sancheti described the GCC shift as an evolution in three phases. In the first model, budgets sit globally, and India teams execute. In the second, local stakeholders co-own delivery but do not carry final authority. In the third, new, independent workstreams emerge inside GCCs with their own budgets and stakeholders, particularly in R&D, engineering-led work, and AI CoEs.

He pointed to healthcare and life sciences, where drug discovery and other R&D functions increasingly run from India centres. He cited industrial and manufacturing customers running engineering work and physical-to-digital IT–OT integration from India.

Sancheti has highlighted a newer category of investment, which is AI centres of excellence for improving developer productivity as organisations move from a traditional software development lifecycle to an AI-driven lifecycle.

For partners, the key change is not the organisation chart. It is where “new work” is born.

In earlier GCC models, India largely executed what was designed elsewhere. In the newer model, India-based leaders increasingly shape what gets built and how quickly it moves from prototype to production.

A co-delivery model built for the new buyer

Tredence’s response to this shift is an embedded GCC co-delivery model built around three pods: co-build, co-skill and co-anchor.

The logic is to move from project execution to capability anchoring. Co-build aligns with rapid solution development and early production outcomes. Co-skill focuses on building internal capability so that GCC teams do not remain dependent on external talent. Co-anchor creates long-term presence around repeatable workstreams where GCC leaders need predictable delivery at scale.

Sancheti argued that this is where many GCC initiatives get stuck: they may have mandates, but they struggle to access consistent talent quality at scale, even after setting up India operations. In that gap, partners become less like downstream vendors and more like execution infrastructure.

He also made a second point that matters for the channel ecosystem. GCCs increasingly operate as distinct contracting entities, often requiring separate agreements independent of global frameworks. That, in practice, reinforces their status as separate buying units rather than extensions of global procurement alone.

Tredence said it has opened eight to ten GCC accounts over the past two years, positioning GCC engagement as both an extension of global relationships and a pathway into newer India-based buying centres.

Why hyperscalers still need partners to ship outcomes

As hyperscalers expand their own consulting and AI capabilities, a common concern across the ecosystem is whether partners risk being edged out as platforms move further up the value chain.

Sancheti framed this as a misconception about how AI solutions are built and scaled in practice.

He said customers today expect hyperscalers and data platforms to articulate industry relevance, not just product features. In response, platform providers have invested in sector-specific frameworks and solution blueprints across industries such as energy, financial services, insurance, retail and CPG.

However, these are starting points, not production systems.

Sancheti said these frameworks are not designed to run at scale inside real enterprises, particularly across fragmented environments such as GCC-led organisations. As a result, hyperscalers depend on partners to take initial concepts and convert them into deployable solutions by adding domain-specific KPIs, agents, notebooks and operational accelerators.

That division of responsibility defines how most large AI programmes actually get executed. Hyperscalers remain product- and platform-focused, with services teams geared towards initial deployments, specialist intervention and high-priority issues. They are not structured to scale implementation or run long-tail execution across multiple business units.

Partners bring industry context and process depth, taking ownership of implementation, expansion and ongoing operations. In many cases, Sancheti said, Tredence delivers this work under hyperscaler contracts itself, reinforcing a complementary rather than competitive model.

This dynamic becomes pronounced as GCCs take on greater ownership of AI mandates. As platforms build GCC-focused sales teams, Tredence aligns its own GCC coverage with both global and regional partner GTM teams, enabling joint co-selling into India-based buying centres.

The moat, Sancheti argued, sits not in owning platforms or frameworks, but in repeatedly operationalising them at scale. As AI moves from experimentation into production inside GCCs, execution capability, not platform control, determines which partners remain relevant.

Partnerships as consumption economics, not badge collection

Tredence framed its partner strategy less as a brand stack and more as a consumption model.

Sancheti argued that one of the biggest opportunities in the ecosystem sits inside committed business already sold by hyperscalers, reflected in metrics such as remaining performance obligations, and the downstream work needed to turn those commitments into real platform usage through AI solutions that drive compute and consumption.

He described three partnership vectors.

The first is platform partnerships with hyperscalers and data and AI providers. The second is a “native AI” ecosystem built around foundation model partners such as Anthropic and OpenAI. The third is partnerships with niche, industry-specific players, including domain specialists in areas such as martech and vertical specialists across sectors like retail, CPG and industrial manufacturing.

The intent, he said, is not to carry partnerships “on paper” but to be meaningful partners, ideally among the top three for those organisations, by acting as a value enabler for customers and a driver of faster consumption.

He said Tredence runs joint initiatives, including partner days, hackathons and programmes like AI Foundry designed to pull partners into use case discovery and rapid prototyping.

In a GCC-led world, this becomes more than ecosystem messaging. It becomes a GTM requirement. If GCCs are becoming new buying centres for AI, partners need repeatable motions that translate platform capabilities into production outcomes quickly, because that is what drives consumption and expansion.

Diversifying without breaking gravity

Tredence acknowledged its exposure to the US market, with 75 percent of revenue coming from the region, while much of its workforce, almost 80 percent, sits in India. Sancheti said the dependence is real, but not unique, reflecting where global AI services spending remains concentrated.

At the same time, he framed diversification as an active strategy.

“We have already made progress on diversifying our regional presence. For instance, we recently hired a dedicated Chief Business Officer for Europe, which shows we are serious about expanding our Europe business,” said Sancheti.

He added, “While the dependency exists, we are not seeing any reduction in customer conversations. In fact, demand has grown over the past one to one-and-a-half years. We believe we are well-positioned with our current capabilities, while continuing to execute on our diversification strategy.”

Tredence aims to move towards a 70:30 regional mix.

In other words, the de-risking path does not necessarily require a clean break from the US-heavy portfolio. It requires learning where the next decision-makers sit and building models that sell into them.

IPO ambition, and the architecture behind it

Sancheti said Tredence intends to pursue an IPO, with a clear focus on profitable, sustainable expansion while reducing portfolio concentration risks.

He pointed to the company’s track record of consistent growth over 13 years, including an eightfold increase in revenue over the past eight years and sustained annual growth of 40–42 percent. The company is targeting a billion-dollar revenue milestone by 2030, with interim goals such as crossing the half-billion mark.

According to Sancheti, in India, the company’s commitment is strong and twofold. First, India will remain central to the talent strategy. Given the strength and depth of AI talent here, Tredence believes India will continue to be a leader in this space for the next 10 to 15 years.

“While we are also building nearshore capabilities in regions like Latin America, Europe, and the Middle East, India will remain our primary delivery hub. We are continuing to expand our delivery centres across the country,” said Sancheti.

On the strategic path forward, Sancheti added, “We intend to pursue an IPO. At the same time, we are open to selective acquisitions to build regional capabilities or deepen expertise in specific industries. For example, we recently acquired Further Advisory, a management consulting firm.”

However, Sancheti clarified that Tredence does not see itself being acquired by a large IT company.