Episode 117 · Enterprise · 44 min

The transformation engine transforms itself

EY's Global Delivery Services sells end-to-end transformation to 2,500-plus clients, and the argument here is that the pitch only earns trust because the century-old firm put itself through the same wringer first — about 95% of its own estate into the cloud in roughly three years. Two harder claims follow: compliance belongs on the first slide rather than in the test phase, and AI will not displace people. People with better AI will.

RR
Ragu Rajaram
Leader, Cloud Consulting Services & Financial Services, EY Global Delivery Services (EY GDS) · with Vishal Krishna
The transformation engine transforms itself — episode thumbnail
44:00
Said in this episode
▶ 1:34
2,500+
Clients EY GDS serves worldwide
Described as 'close to 2,500 plus' clients across every sector — financial services, media and telecom, healthcare, mining — wherever EY has offices.
▶ 6:40
500+
Clients in the cloud consulting book
The cloud consulting practice he leads serves 'close to 500 plus' of those clients, having run what he calls a tremendous number of cloud transformations.
▶ 16:54
~95%
Share of EY itself running in cloud
A 100-year-old tax and accounting firm that he says moved itself in roughly three years; the audio garbles the phrasing around the number, so treat it as an as-stated figure.
▶ 36:30
~400,000
EY people to be trained on foundational AI
Every person in the firm, at any level, is to go through foundational AI courses and certification; he says the programme is well under way and the firm has invested billions of dollars in the surrounding infrastructure.
▶ 18:48
1/4th
The old India cost pitch
Twenty years ago the GCC story was roughly one-fourth of onshore cost, varying by geography; his argument is that the sale is now capability and scale rather than arbitrage.
The brief

The argument in sixty seconds

Rajaram's claim is that a transformation business only sells if it has already survived its own transformation. EY is a century-old tax and accounting house that pivoted hard into consulting around 2011 and now, he says, runs roughly 95% of itself in the cloud after about three years of work — which is the licence it uses to walk 2,500-plus clients from the boardroom to the server room. Everything else follows from that. Compliance and regulatory review get dragged to the far left of a programme instead of being discovered during testing, because the failure mode is not a bug, it is the front page of a newspaper. Cloud has stopped being a storage or a cost conversation and become a tempo one — capex to opex, elasticity bought for a series launch and released the week after, and his blunt image of standing still on a moving treadmill. Banks, hospitals and governments are not slow, only cautious, moving one workload at a time, which is exactly where hybrid came from. And India's offer has flipped: twenty years ago the pitch was a quarter of the cost, now it is that the country is the transformation engine, staffed by problem solvers rather than order takers. The stakes sit in the final third. EY is putting close to 400,000 people through foundational AI training, because in his framing AI never competes with a human — humans compete with humans, and the one holding the better AI wins.

Worth your time if you are

CIOs sequencing a hybrid-cloud move under a regulator's eye
Consulting and GCC leaders selling one programme to four different C-suites
HR and L&D heads planning firm-wide AI reskilling
Engineers in India's services industry wondering what the next decade rewards
Graduates choosing between a job and a career path
Episode map

Where the conversation travels

Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.

01Cold open: the running order 0:00 Vishal sets the agenda in ninety seconds — AI and generative AI, cloud, security, compliance, culture and health and wellness — and Rajaram states his brief: cloud consulting for EY plus the financial services sector for Global Delivery Services. 02From the boardroom to the server room 1:19 GDS is described as a syndicate inside EY serving 2,500-plus clients across every sector, built out of a deliberate 2011 pivot from tax and accounting into consulting — where consulting means handholding through execution, not handing over a blueprint. 03Compliance, dragged to the far left 4:39 Technology teams chasing a differentiated proposition skip the compliance question, and the penalty is not a defect but the front page of a newspaper — so regulatory and consent questions are made a forethought, baked into design rather than surfaced in testing. 04Cloud is a tempo problem, not a cost one 6:35 With 500-plus cloud clients, Rajaram argues cloud is the bedrock of any modernisation — greenfield goes cloud-first, but the real work is uncluttering decades of legacy — and the payoff is agility and reach, not just the cost line, because standing still on a treadmill means falling off. 05Spotify can, a bank can't 9:40 Public, private and hybrid split by how much regulated data you sit on: streaming firms can run wholly public and scale elastically for a launch month then descale, while banks, healthcare and government move one workload at a time — cautious rather than slow, horses for courses. 06Four chairs, four different agendas 14:00 The CEO wants advancement, the CFO wants optimal cost, the CIO and CTO want something cool to sell, the security chief wants no new risk — so the same programme has to be argued four different ways, because a stakeholder left out of the business case sinks the journey. 07The century-old firm moved first 16:06 Asked about cloud-native clients, Rajaram pivots to EY's own transformation — a 100-year-old accounting firm now running about 95% in cloud after roughly three years — and notes that inherently technical companies mostly do their own cloud work and come only when scale demands help. 08From a quarter of the cost to the engine 18:25 The GCC story began twenty years ago at roughly a quarter of onshore cost; today, he argues, India is in a golden age as the world's transformation engine, and the workforce has gone from order takers to problem solvers who negotiate as stakeholders. 09Culture first, and upskilling beat salary 21:00 Growing from a small team meant refusing to be everything to everybody and guarding culture through hiring and acquisition — and an internal survey surprised him by ranking the chance to work on advanced technology above salary, with work-life balance third. 10Europe writes the AI rules first 23:35 Most of the work comes from the Americas, but delivery hubs in Poland and Malaga serve a Europe with far more regulation than the US — the first to legislate on AI — which turns responsible-AI onboarding into a consulting line of its own. 11APAC's spread, and QR-code envy 26:06 APAC is even more culturally scattered than Europe, and the arbitrage of a follow-the-sun firm is carrying a lesson from Australia into a European client — while American visitors arrive every week and are floored by UPI, QR-code payments and ten-minute delivery the West never built. 12Fitness as an operating discipline 30:12 Running a business across Australian, European and American clocks, he refuses to touch the phone on waking and spends an hour and a half in the gym instead — then argues the firm's fitness carnivals, 10K runs and WhatsApp peer pressure exist because only healthy, happy people are productive. 13Humans versus humans with better AI 34:45 His answer to the fear in the room: AI will not displace a person, the contest stays human against human, decided by who adopts faster — and generative AI landed in a single year only because it was democratised down to grandparents and children. 14Foundations, free-field years, career paths 37:30 Close to 400,000 people are being put through foundational AI training on an internal, firewalled LLM platform so company documents never leave the building; campus hires get two free-field years across cyber, data, digital and cloud before choosing — a career path, not a job. 15Grit, music and the fire in the belly 40:55 He closes on a humble background and the book Grit, on comparing himself only to his previous self, on the evergreen film music he finds meditative, and on a wife and three children who are the reason the balance question matters at all.
Takeaways

Ideas to carry out of this hour

01

Compliance belongs on the first slide, not in the test phase

The instinct in a differentiated technology programme is to chase the niche proposition and deal with regulation later, which means discovering during testing that a consent or data rule was never designed for. Rajaram's rule is the reverse: compliance and regulatory sit as a forethought, pulled to the far left of the programme and baked into the design and engineering model. The reason he gives is not legal, it is reputational — the last thing a client wants is to be on the front page of a newspaper for the wrong reason, and customer data is the raw material of every insight anyone is trying to build.

02

Cloud stopped being storage and became tempo

Most people still think of cloud as the place their phone photos go, and most boards still approve it as a cost line. He argues the cost case is only balanced, and the real returns are agility, innovation and market reach — the shift from capex to opex means you rent what you need when you need it and can break barriers without a procurement cycle. The image he keeps returning to is a treadmill: the industry is already moving, and an organisation that stands still on it does not stay level, it slips and falls off.

03

Regulated industries are sequencing, not stalling

A streaming company can run entirely on public cloud because it does not sit on regulated data; a bank cannot be compared to it. That gap, he says, is where hybrid cloud was born, and it is why every hyperscaler now offers a private-cloud footprint inside the client's own data centre. Financial services, healthcare under HIPAA and government are not laggards — they are making reasonable judgments, moving one workload at a time, horses for the courses.

04

The firm selling transformation had to survive its own

Asked why a cloud-native unicorn would ever hire a hundred-year-old accounting firm, Rajaram answers with EY's own migration: a century-old tax and accounting organisation that now, by his account, runs roughly 95% in cloud after about three years. The claim is that the credibility is earned rather than asserted — the learnings from doing it to themselves are what they now carry into client programmes. He is equally candid about the limit: inherently technical companies know the cloud story and mostly do it themselves, coming to EY only when scale or geography outruns them.

05

One programme, four buyers, four business cases

A transformation touches every part of the value chain, so it has to be sold four ways at once: advancement to the CEO, optimal cost and on-time delivery to the CFO, something genuinely new to release to the CIO and CTO, and risk containment to the security chief. Rajaram's point is that this is a sequencing problem, not a pitch problem — if one stakeholder is not party to the case at the start, the journey becomes difficult for everybody. The unlock is early visible success, because once there is a success story every function wants to be inside it.

06

India's pitch flipped from a quarter of the cost to the engine room

Twenty years ago the global capability centre existed on arbitrage — roughly a quarter of onshore cost, and every GCC told the same story. He argues the sale has moved to capacity, capability and scale at an optimal price, with India as the transformation engine of the last two decades, having held its ground while other geographies tried to substitute for it through forex and geopolitical shifts. The internal change matters more than the external one: teams that were told what to do have become teams that negotiate as equal stakeholders and solve the client's problem alongside them.

07

Their own people ranked upskilling above salary

EY GDS surveyed employees on the three things that mattered most and Rajaram expected pay to top the list. It came second. First was the opportunity to upskill and work on advanced technology; work-life balance came third. That result underwrites the training machine described later — close to 400,000 people across the firm being put through foundational AI courses regardless of level, plus domain certifications in banking and insurance and a sponsored MBA — and it reframes retention as a curriculum problem rather than a compensation one.

08

AI does not compete with you; the person using it does

His answer to the newspaper-induced fear is deliberately narrow: AI is not going to displace a human, because the competition remains between humans — decided by who has the better AI and adopts it faster. The reason generative AI made its dent in roughly a year, when AI itself is decades old, is democratisation: a grandparent, a parent and a child can all now use it in some capacity. Which is why the firm's answer is training everybody rather than hiring a specialist team, and why the internal work runs on a firewalled LLM platform so company documents never end up in a public chatbot.

The numbers, drawn

What the episode measures

Every figure below was said on air — timestamps included, caveats kept.

Conversation share

portion of the hour spent on each theme
GCCs & services · 20%AI & machine learning · 18%Data & digitisation · 16%Hiring & talent · 13%Regulation & policy · 11%Leadership & org · 10%
GCCs & services20%
AI & machine learning18%
Data & digitisation16%
Hiring & talent13%
Regulation & policy11%
Leadership & org10%
Computed from the chapter map of this episode.

Where the cloud book sits inside GDS

clients
GDS clients, all ser2,500Cloud consulting cli500
Both figures exactly as stated on air, each given as 'close to … plus' — so roughly one client in five sits inside the cloud consulting practice he leads.▶ 6:40
Worth keeping

Lines that stay

Whatever they did was the right thing at that point in time — and by the way, whatever you do today is legacy tomorrow.

— Ragu Rajaram ▶ 7:21

It's like standing on a treadmill. Some of them just stand on the treadmill — they're going to slip and fall. You've got to be at the pace of the treadmill.

— Ragu Rajaram ▶ 8:07

AI is not going to displace a human. The competition is still going to be between humans and humans — but who's got the better AI.

— Ragu Rajaram ▶ 35:24

We need healthy people, we need happy people — and they are the only ones who will be productive.

— Ragu Rajaram ▶ 33:07

That's the difference between giving somebody a job and giving somebody a career path. A job is a transaction.

— Ragu Rajaram ▶ 40:43
Clips that travel

Short on time? Start here

CIOs and risk leads scoping a transformation programme

Compliance as a forethought, not a test-phase surprise

The cleanest statement of the shift-left argument — consent, customer data and why the real downside is a newspaper front page, not a defect.

4:39 → 6:35 · 2 min ▶ Watch clip
Boards still approving cloud as a cost line

The treadmill, and the capex-to-opex flip

Why cloud is agility rather than savings, why every greenfield build is cloud-native, and the image that reframes standing still as falling behind.

6:35 → 9:40 · 3 min ▶ Watch clip
GCC and services leaders repositioning an India centre

From cost arbitrage to the transformation engine

The quarter-of-the-cost origin story, India's golden age claim, and the shift from order takers to problem solvers with swagger.

18:25 → 21:00 · 3 min ▶ Watch clip
Fintech and commerce operators explaining India abroad

What American visitors notice about India

UPI and QR payments, ten-minute delivery and scale as the things the West studies but cannot easily copy — and what India still learns from the West.

27:23 → 30:12 · 3 min ▶ Watch clip
Mid-career professionals reading the AI headlines with dread

Humans versus humans with better AI

The displacement question answered directly, why generative AI landed in a single year, and the 400,000-person training response.

34:45 → 37:30 · 3 min ▶ Watch clip
Glossary

The jargon, unpacked

GDS (Global Delivery Services)
EY's global delivery arm — described here as a syndicate inside the firm that runs strategy through to end-to-end execution for clients wherever EY operates.
GCC (Global Capability Centre)
An offshore centre a multinational runs itself rather than outsourcing; India's GCC pitch began as cost arbitrage and, he argues, has become capability and scale.
Hybrid cloud
Splitting workloads between public cloud and private infrastructure so regulated data stays where a regulator can see it — born, in his telling, from the gap between what a streaming firm and a bank can each put on public cloud.
Capex to opex
Moving from buying and owning infrastructure as capital expenditure to renting it as an operating cost, paying only for what you use as you use it.
Shift left
Moving a step — here compliance and regulatory review — to the earliest stage of a programme rather than the end, so it shapes the design instead of blocking the test phase.
Autoscaling / elasticity
Adding cloud capacity ahead of a demand spike, such as a new series launch, and releasing it once traffic falls — economically impossible on owned infrastructure.
EU AI Act
Europe's AI legislation, cited as the first of its kind, which pushes organisations adopting AI in Europe towards documented responsible-AI practice.
Responsible AI
Governance, transparency and risk controls applied to AI systems so their use stands up to regulators — for EY, a consulting line created by Europe's rules.
Connections

If this resonated, go here next

Full transcript

The whole conversation, searchable

172 segments

Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.