Chamath’s $135 Million Software Factory Puts Europe’s IT Giants on Notice

Chamath Palihapitiya raised $135 million in June to build software that could put IT consulting giants out of business, and he named Accenture, Infosys and Wipro as the targets. Capgemini and Atos never came up.

Between them, Europe’s two biggest IT services firms employ just over 484,000 people. Both had already told investors, in their own words, that AI is squeezing the business model 8090 wants to replace.

A $135 Million Bet Aimed at Accenture’s Business Model

8090 Labs closed its Series A on June 29. Salesforce Ventures led the round, and the rest of the cap table read like a guest list: WndrCo, Craft Ventures, The Production Board and Launch, the venture firms run by Palihapitiya’s fellow hosts on the All-In podcast. Angel checks came from Palo Alto Networks CEO Nikesh Arora and Quora CEO Adam D’Angelo.

Palihapitiya founded 8090 Labs in January 2024. The raise came with a personal move too: he stepped off the board and into the CEO chair, his first full time operating role since he left Facebook in 2011.

The product is called Software Factory, and the pitch is that it can put people and AI agents in one workspace to build enterprise software with the audit trails and oversight regulated industries demand. “AI can write code,” Palihapitiya said in the round’s announcement. “The hard part of enterprise software is keeping fifty agents and a hundred engineers changing the same complex system every week without it pulling apart.”

The target list is specific: healthcare, insurance, life sciences, aerospace, energy, manufacturing, financial services and the U.S. government, the exact regulated sectors that built the market for Accenture, Infosys and Wipro over decades of staffed engagements. Tech Times reported that Software Factory prices self-serve access at $200 a user per month plus token based usage, with fully managed enterprise deployments starting at $1 million a year. The AI coding market is already crowded, from GitHub Copilot to Cursor to Meta’s newest coding-agent price and performance release, and 8090 is trying to sit a layer above all of them by selling governance rather than raw code generation.

The Evidence Behind the Sales Pitch

8090 backs its pitch with customer results it says prove the model works in production, not just in demos. The company is upfront that these are its own figures, not independently checked.

• Healthcare billing engine: reverse-engineered 18 million lines of COBOL and Assembly into more than 300,000 plain English rules in 40 days.
• Health insurer claims: cut claims routed to a pay-per-catch vendor by more than 80%, avoiding over $20 million across four years.
• Life sciences diagnostic: cut time to market for a new diagnostic from five years to four.
• Manufacturer parts validation: brought more than 10,000 parts under real-time validation with automatic approvals reaching over 1,000 users.

That healthcare case is the one 8090 leans on hardest. The company says it reverse-engineered 18 million lines of COBOL behind a billing engine most consulting firms would have billed years to touch.

Europe’s Consulting Giants Never Got Named

Every write-up of the 8090 raise reached for the same three names: Accenture, Infosys, Wipro. None reached for Capgemini or Atos, the French firms that built the exact same staffed, billable-hour model on European soil and now employ close to half a million people doing it.

That is not a small omission. Capgemini alone reported 421,000 employees as of March 2026. Atos closed out 2025 with 63,193. Both are pure IT services businesses selling custom software, application maintenance and outsourced delivery, the precise work Software Factory is designed to replace. Both are also publicly traded, which means both have already had to explain, on the record, what AI is doing to their numbers.

Companies selling specialized software and consulting services have already been caught in a broader selloff tied to AI exposure, separate from anything 8090 has done specifically. The pressure predates this one raise. It just got a louder headline.

The Bench Problem Capgemini Already Flagged

Capgemini’s own Q1 2026 numbers look strong on the surface. Revenue hit €5.94 billion, up 11% at constant currency, and generative and agentic AI bookings topped 11% of group orders for the quarter. Aiman Ezzat, the company’s chief executive, said the results validate “our cloud and AI strategy, delivering strong underlying growth, in line with Q4 2025 and outperforming most peers in the market.”

Look closer and the picture splits. Applications and Technology, the segment closest to custom software development and maintenance and Capgemini’s largest at 63% of group revenue, grew a comparatively modest 4.8% at constant currency. Operations and Engineering, boosted by the recently acquired WNS and Cloud4C units, grew 25.2% and did most of the heavy lifting behind the headline number. Much of Capgemini’s 421,000-strong headcount growth traces to those same acquisitions rather than fresh hiring.

Capgemini told analysts on the same call that bench accumulation in Continental Europe, meaning consultants sitting between assignments while still drawing a paycheck, was weighing on margins, even with utilization holding at 78% in Applications and Technology and 70% in Strategy and Transformation. Full year 2025 told a similar story before AI became the headline explanation: revenue grew just 1.7% to €22.5 billion, while net income fell 4.2% to €1.6 billion. Ezzat has separately said the company is “clearly pivoting” to push AI adoption harder in 2026, after acknowledging demand had been subdued in some markets the year before.

Atos Bet Its Turnaround on the Same Technology Threatening It

Atos arrives at this moment from a worse starting position. Its “Genesis” restructuring plan cut headcount by 19% to 63,193 employees, after a planned sale of its legacy IT business fell through and forced the company back to the drawing board. Full year 2025 revenue landed just above target at €8.001 billion, but the core Atos business unit declined 16.2% organically to €6.96 billion, even after a notable cybersecurity contract win with the European Commission.

Atos is responding by going all in on the same technology squeezing it. The company’s own Data and AI unit is growing from 2,000 to 10,000 employees by 2028, with a target of certifying its entire workforce in AI by 2026. In June, Atos said it would deploy Microsoft’s agentic AI tools, including Microsoft 365 Copilot, across its global workforce, extending the same security and compliance workflows it runs for human staff to its new digital one.

The company says it has completed 88% of its three-year savings target in under a year. It is targeting positive organic revenue growth in 2026, 5 to 7% annual growth by 2028, a 10% operating margin, and an investment-grade credit rating. Every one of those targets assumes AI makes Atos more competitive before it makes Atos’s own delivery model obsolete.

Is 8090 Really Breaking a Fifty-Year Cycle?

Not entirely, but Palihapitiya says AI ends a pattern that built the entire IT services industry: a company writes its own software, then outsources it to a vendor, then offshores the maintenance while costs climb and quality slips. That exact pattern is how Capgemini and Atos got to nearly half a million employees between them. Software Factory is betting AI removes the need for every step of it.

For 50 years, we’ve watched the same cycle repeat. A company initially writes their own software, then outsources it to a commercial vendor, then offshores the maintenance of that system, all while costs keep rising and quality keeps falling. We built Software Factory to solve this problem.

Palihapitiya wrote that in a separate statement reported by Tech Times, laying out the logic behind Software Factory’s design. It is a clean argument. It is also, notably, an argument that assumes the cycle breaks everywhere at once, not just in the U.S. market 8090 currently serves.

Can GDPR Slow Down America’s Software Factories?

Possibly, for a while. Europe’s GDPR rules and incoming AI Act obligations mean regulated buyers in healthcare, finance and government, 8090’s exact target list, need proof that code and sensitive data never leave EU jurisdiction. Guides built for European developers already treat data residency as a serious selection filter for U.S. coding tools, citing the ongoing legal uncertainty around EU-U.S. data transfers under GDPR’s cross-border rules. 8090 has said nothing publicly about EU hosting or data residency.

That gap will not last forever. Capgemini and Atos already sell sovereignty and compliance as a feature, not an afterthought. Capgemini’s own results statement pointed to “growing momentum” in demand for sovereign technology, and Atos has built its AI rollout around the same compliance workflows it already runs for human staff. If a Europe-based rival, or a European arm of 8090 itself, closes that compliance gap, the advantage evaporates fast.

Capgemini’s next quarterly results are due this summer. Atos keeps executing Genesis through 2028. Neither has said 8090’s name out loud yet.