The News

Infosys (Reference - Indiainfoline - 18th Nov) has announced a new AI-First model for Global Capability Centres (GCCs), with the goal of transforming them into AI-driven innovation hubs rather than traditional back-office units.

According to the announcement:

• The model is built on Infosys’ experience managing over 100 GCC projects, including work with Lufthansa Systems, zooplus, and Danske Bank.

• Infosys will support enterprises across the full GCC lifecycle: strategy, site selection, recruitment, operational launch, and long-term management.

• The framework integrates AI tools and platforms into GCC operations to improve process efficiency and enable AI-powered innovation.

• It draws on several Infosys platforms:

  • Agentic Foundry – to build reliable AI agents.

  • EdgeVerve AI Next – to run AI at enterprise scale.

  • Infosys Topaz – to embed AI-powered services across GCC operations.

• Infosys cites a recent engagement with Lufthansa Systems, where they helped set up a dedicated GCC focused on advanced aviation IT products using generative AI for safety, efficiency, and customer experience.

• The model addresses talent needs by using Infosys Springboard digital learning platform and its corporate university to train employees and maintain a pipeline of skilled professionals.

• Infosys offers multiple GCC engagement models, including:

  • Build-operate-transfer (BOT)

  • Assisted builds

  • Joint ventures

  • Partner-hosted arrangements

The objective is to help companies treat GCCs as strategic engines of innovation and growth, not just low-cost delivery centres.

Why This Matters Now

If you’ve worked with GCCs before, you know the pattern:

They start with big words like “global hub” and “innovation engine,”
then quietly devolve into ticket queues, BAU dashboards, and SLA firefighting.

Infosys is trying to reverse that gravity.

This announcement matters because it shows where large enterprises are actually moving:

• From scattered AI pilots to structural AI programmes

• From “let’s try a POC in one department” to “let’s re-architect the way our global centres operate”

• From GCCs as cost arbitrage to GCCs as AI-first operating systems

For AI startups and builders, this is a signal:

The next phase of demand won’t just be departmental copilots.
It will be GCC-level transformation where entire regions are retooled around generative AI and automation.

If you want to plug into that wave, you need to understand how these centres are being reimagined.

What Is Being Built or Changed

GCCs as strategic AI assets

The pitch is simple:
GCCs are no longer just execution arms.
They are being reframed as strategic assets that drive efficiency, agility, and growth.

Infosys is giving large enterprises a model that says:

• Don’t set up a GCC just to cut cost.

• Set it up to experiment, productise and scale AI across the organisation.

A full-lifecycle AI operating model

This isn’t just a one-time consulting engagement.
Infosys is offering end-to-end involvement:

• Strategy planning – what the GCC is meant to own.

• Site selection – where to locate it, with talent and cost in mind.

• Recruitment – building teams with AI, engineering, and domain skills.

• Operational launch – getting the centre running with AI baked in from day one.

• Long-term management – ongoing optimisation, scaling, and governance.

The key detail: AI isn’t bolted on afterwards.
It’s designed into the GCC’s operating model from the start.

Platformised AI: Agentic Foundry, EdgeVerve AI Next, Infosys Topaz

The model doesn’t exist in a vacuum.
It’s anchored on three platform pillars:

• Agentic Foundry – building reliable AI agents that can orchestrate tasks across systems.

• EdgeVerve AI Next – a platform to run AI at enterprise scale, across complex landscapes.

• Infosys Topaz – AI-powered services threaded through GCC processes and operations.

Together, they turn GCCs into places where:

• Agents handle routine workflows

• AI supports decision-making

• Automation reduces manual overhead

• AI products and solutions can be built and reused across business lines

Talent as a first-class concern

This is one of the few enterprise AI narratives that explicitly calls out talent as part of the model, not an afterthought.

Infosys plans to use:

• Springboard – its digital learning platform

• its corporate university – to keep people upskilled

This means the AI-First GCC model isn’t just about tooling — it’s about:

• who you hire

• How you train them

• How you keep the bench relevant as AI tech evolves

Flexible engagement models

Not every enterprise wants the same GCC structure.
Infosys offers:

• BOT (build-operate-transfer) – Infosys sets up and runs the GCC, then hands it over.

• Assisted builds – the enterprise leads, Infosys co-pilots.

• Joint ventures – shared ownership.

• Partner-hosted – Infosys or partners host environments, with clients leveraging capacity.

This flexibility is important because it allows enterprises to move at different speeds and risk tolerances, while still aligning to an AI-first strategy.

The BitByBharat View

I’ve seen enough “new operating model” decks to know that most never reach the infrastructure.

They stay at the slogan layer.

What’s different here is that Infosys is pinning its model to:

• Concrete platforms

• A real client (Lufthansa Systems) using generative AI for advanced aviation IT

• Explicit lifecycle involvement

• Real talent pipelines

In other words: it’s closer to a playbook than a positioning slide.

If you’re an AI founder or product builder, this matters because:

• GCCs are where a lot of quiet decision-making happens about tools, platforms, and vendors.

• If those centres are being rebranded as AI innovation hubs, the kind of tools they seek will change.

• They’ll be looking not just for pilots, but for products that plug into their AI-first architecture.

The pattern is clear:

• Cloud was the last wave.

• AI-first GCCs will be part of this wave.

If you stay stuck selling into isolated teams, you’ll miss the bigger consolidation happening at the organisation’s centre of gravity.

The Dual Edge (Correction vs Opportunity)

Correction

Many AI startups still think in terms of:

• One team

• One use case

• One POC

This model says:
Enterprises are starting to think in systems again, not isolated experiments.

If you ignore GCCs and centre-level transformation, you risk being:

• A niche add-on

• Bypassed in standardisation

• Excluded from long-term stack decisions

Opportunity

On the flip side, if GCCs become AI-first innovation hubs, they’ll need:

• Agent frameworks that integrate with their existing systems

• Governance, observability, and monitoring tools for AI workflows

• Domain-specific copilots that plug into their platforms

• Custom evaluation frameworks for AI inside regulated environments

• Tools that help them build once, reuse many times across business units

There is a lot of room for:

• Focussed vertical solutions

• AI infra tools that sit above platform APIs

• Specialist companies that understand both GCC constraints and AI potential

This is where thoughtful builders can win.

Implications (Founders, Engineers, Investors)

For Founders

If your product is aimed at the enterprise:

• Start mapping where GCCs sit in your target organisations.

• Understand whether they are being tasked as innovation hubs.

• Position your product not as a one-off experiment, but as part of the AI-first operating model.

Ask:
Could our product live inside an AI-first GCC blueprint?
If the answer is yes, your messaging and integrations should reflect that.

For Engineers

Look closely at the technology stack implied here:

• Agent orchestration (Agentic Foundry)

• Enterprise AI platforms (EdgeVerve AI Next)

• AI services woven across operations (Topaz)

• Integration with legacy systems in aviation, banking, retail, etc.

This is the world you’ll be building into:

• Multi-system, multi-region, heavily regulated

• Automation as a default expectation, not a bonus

• AI living at the centre of operational workflows, not at the edges

Technical leverage will come from understanding how to:

• Design agents that are reliable at scale

• Plug into existing enterprise platforms

• Make your tools observable and governable

• Support GCC-style multi-team, multi-region use

For Investors

This kind of announcement is a directional signal:

• Large enterprises aren’t just “trying” AI anymore; they’re restructuring around it.

• GCCs are a key vehicle for that restructuring.

• Service providers like Infosys are codifying that shift into frameworks and products.

You’ll want to look for:

• Startups that help operationalise AI in GCC-like environments

• Tools that make enterprise AI legible to business stakeholders

• Products that turn GCCs into reusable AI “factories” for the rest of the company

The money will follow the centres that can repeatedly ship AI into production — not just win awards for innovation labs.

Closing Reflection

GCCs used to be treated like distant factories: somewhere offshore, doing necessary work, mostly invisible.

Infosys’ AI-First model is one more sign that this view is outdated.

Those same centres are being invited into the core of how companies innovate, automate, and serve customers — with AI acting as the accelerant.

As a builder, it might be worth revisiting your own mental map:

• Where does AI “live” in the organisations you care about?

• Are you still thinking in pilot projects, or are you designing for centres that will own AI at scale?

Because as soon as GCCs become true innovation hubs, they’ll need tools, platforms, and partners that think beyond demos.

If you’re building one of those, this is a good time to align your roadmap.

Reference - Indiainfoline - 18th Nov