Wednesday, October 7, 2026Ground Truth MagazineSubscribe
Digital Transformation

Walking the graveyard where the castles are buried: what GE’s Predix teaches about capital

GE spent more than $7 billion on Predix and still could not make it stand. As Algorizz builds its own manufacturing operating system, its founder argues the root cause was money.

AS
By Published · 4 min read
ShareinWA
Ruins of an old stone castle standing in an open field
Big budgets set big promises, and the promise ends up writing the strategy. Photo: Yana Tes / Unsplash

The castle

Algorizz has been prototyping a manufacturing operating system. While researching the space, founder and CEO Rajneesh Mittal kept returning to one cautionary tale: GE’s Predix[1].

Predix cost GE about $7 billion[1]. The company built a grand vision and sold it, first to itself. GE Digital’s then leader Bill Ruh forecast $15 billion in software revenue by 2020, and Jeff Immelt[3] had already promised GE would be a top ten software company by then.

On paper it was a textbook digital transformation story and a visionary product. Platform as a service, a developer marketplace, digital twins and predictive analytics. A billion-dollar campus in San Ramon. Hundreds of engineers hired from top tech companies, a chief digital officer recruited from Cisco, and consultants from the most prestigious firms.

December 2018

By December 2018 the business was being spun out with about a billion dollars in revenue and no CEO. Ruh left. Three years later, GE announced it would split into separate companies[2].

What was the bug?

“

What was the bug? Money.

Mittal’s diagnosis is blunt. A company that spends seven billion dollars cannot then settle for a modest business. The spend sets the promise, and once the promise is set, it writes the strategy. If you have booked Madison Square Garden, you have to fill it. Not everyone is Taylor Swift.

The Leadership OS for the AI Era
Also readThe Leadership OS for the AI Era

Not an argument against capital

This is not an argument against capital. Some things cannot be built small. A semiconductor fab cannot be prototyped, and a frontier AI model cannot be trained in a garage and scaled up later. TCS[5] is putting billions into gigawatt-scale data centre capacity[4], and there is no half-sized version of that bet that would teach it anything.

The question Mittal asks is whether the money bought something that could not have been bought for a fraction of the price with the same result. Seven billion dollars buys a campus, pedigrees and consultants. Could the job have been done for 70 million, or 1% of the budget?

“

Capital is not the disease. The recklessness is.

The same pattern appears at a far smaller scale. A startup raises a few million and starts spending as if there is no tomorrow, hiring expensive talent whether it needs it or can afford it. As the line often credited to Warren Buffett goes, if you buy things you do not need, one day you will have to sell things you need.

Brick by brick

Algorizz is tackling the same problem Predix set out to solve, but with a deliberately lean approach, spending only what the work requires. Mittal is candid that frugality is easy without billions in the bank, and that a company of Algorizz’s size is in no position to preach. For him the Predix story is a reminder of the risks to watch for while building a business brick by brick.

References and further reading

  1. GE Digital and the Predix platform. Wikipedia
  2. General Electric: history, restructuring and break-up. Wikipedia
  3. Jeff Immelt. Wikipedia
  4. Data center. Wikipedia
  5. Tata Consultancy Services. Wikipedia
RM
In this story
Rajneesh MittalFounder & CEO, Algorizz

An IIT BHU alumnus with 27 years in the technology industry, Rajneesh has been CTO at Zee Entertainment and Manipal Group, with earlier leadership roles at Vodafone, HP, Reliance and Hughes. He founded Algorizz, a Bengaluru-based AI and technology company.

More from Algorizz Signal