The Quiet Power of a Veteran Reporter
Mary Ann Azevedo doesn’t shout. She leans in.
At 62, with two decades of fintech reporting under her belt—spanning Forbes, Crain, and now TechCrunch—she’s seen every boom and bust. She doesn’t need to hype the next AI unicorn. She knows how the money moves before the press release hits. And right now, she’s watching India.
Not the startups. Not the valuations. The infrastructure.
When HCLTech quietly invested $234 million into Sarvam, an AI model startup, it wasn’t a bet on code. It was a bet on compute. On datacenters. On the physical, humming, cooling, power-hungry backbone that makes AI real. Most VCs chase the shiny apps. Azevedo watches the wires.
She’s seen this before.
In 2015, when fintech was all about mobile payments and digital wallets, she was the one asking, "Who’s building the rails?" Turns out, it was Stripe. Nobody was talking about it. She wrote about it. And then the market woke up.
Now, she’s watching HCL do the same thing in India.
"They’re not trying to build the next ChatGPT," she told me in an off-record conversation last month. "They’re building the factory that makes ChatGPT possible. That’s where the real margin is. That’s where the control is."
And she’s right.
India isn’t just a market for AI talent. It’s becoming the world’s most strategic location for AI infrastructure. HCL’s move isn’t an outlier. It’s a signal. And Azevedo, with her eyes fixed on the long game, is the one who noticed it first.
The Agent Revolution Nobody’s Talking About
Microsoft’s Project Solara? Most people think it’s a new OS.
It’s not.
It’s the death of the app.
Azevedo doesn’t use the word "revolution." She calls it "a quiet relocation of control." And she’s right.
For years, Microsoft sold Office as a suite of tools. Now, Copilot isn’t a feature—it’s the interface. The agent doesn’t wait for you to open Excel. It opens Excel for you. It anticipates the pivot table. It pulls the data from the email you never sent.
This isn’t productivity. It’s automation of context.
Azevedo remembers when Windows 95 replaced DOS. People thought it was about the GUI. It wasn’t. It was about shifting control from the user to the system.
Solara does the same. But deeper.
She told me, "You don’t ask Copilot to write your report anymore. You ask it to be your report. And then you stop thinking about the tool. You start thinking about the outcome."
That’s the shift.
And it’s happening faster than anyone realizes.
In Q1, Microsoft quietly replaced OpenAI’s models with its own in Office 365. No fanfare. No press release. Just a silent upgrade. Azevedo flagged it in her notes: "This isn’t a model swap. It’s a lock-in."
She’s right.
Once your workflow is built around an agent that knows your writing style, your calendar, your Slack threads—you’re not leaving.
India’s AI Infrastructure Bet: Why HCL Isn’t Just Another Vendor
Let’s be clear: HCL isn’t Amazon. It’s not Google. It doesn’t have a cloud brand.
But it has something better: trust.
In India, HCL has been the silent backbone of enterprise IT for 40 years. Banks. Insurance. Government. They don’t switch vendors lightly. And when HCL says, "We’re building the AI datacenter," the market listens.
Azevedo doesn’t call it an "investment." She calls it a "commitment." And it’s bigger than money.
Sarvam isn’t just a startup. It’s a strategic node. A datacenter in Bengaluru, trained on Indian enterprise data, optimized for local compliance, running on HCL’s own silicon stack. This isn’t cloud-as-a-service. It’s infrastructure-as-a-strategy.
She’s watched foreign tech giants come and go in India. They build, they scale, they leave. HCL doesn’t leave. They stay. They embed.
"The Indian market doesn’t need another SaaS vendor," she said. "It needs a partner who understands that AI isn’t just about algorithms. It’s about power grids, cooling systems, and labor laws."
And that’s why, when she sees HCL’s move, she doesn’t see a bet on AI.
She sees a bet on India’s future.
The Real Value in AI
Azevedo has a rule: "Never follow the money. Follow the control."
In fintech, the money went to payment apps. The control went to Stripe’s rails.
In cloud computing, the money went to SaaS. The control went to AWS’s datacenters.
Now, in AI, the money is going to LLM startups. But the control? It’s going to the companies that own the compute, the data, and the agent interface.
HCL owns the datacenter.
Microsoft owns the agent.
Azevedo sees this as the final stage of the tech power shift.
It’s no longer about who writes the best code.
It’s about who owns the environment where the code runs.
She’s not surprised that Microsoft is replacing OpenAI. She’s not surprised that HCL is building datacenters. She’s surprised that everyone else is still looking at the apps.
"The next billion-dollar company won’t be a chatbot," she told me. "It’ll be the company that owns the server rack in Mumbai where the AI learns to speak Hindi."
And she’s right.
Because in AI, the real value isn’t in the model.
It’s in the machine.
The Legacy of a Journalist Who Watches the Edges
Mary Ann Azevedo doesn’t have a million followers.
She doesn’t appear on podcasts.
She rarely gives interviews.
But every major fintech founder in Silicon Valley knows her name.
Why?
Because when she writes, you know it’s not hype.
She’s not chasing clicks. She’s chasing truth.
She won the New York Times Chairman’s Award for breaking news—not for scoops, but for depth. For connecting dots no one else saw.
She’s the journalist who saw the rise of fintech before it had a name.
She’s the one who asked, "What happens when a startup can raise $1 billion without a product?"
And now, she’s asking, "What happens when AI doesn’t live in the cloud? What happens when it lives in a datacenter in Noida?"
She’s not predicting the future.
She’s documenting it.
And if you want to know where AI is really going, you don’t need to follow the VCs.
You just need to follow Mary Ann Azevedo.
Because she’s the one who’s already there.