← All digests

India AI DigestJune 25, 2026

India AI Digest — Thursday, June 25, 2026

Vishal Sikka, Infosys's former CEO, comes out of stealth with Hang Ten Systems and a $32M seed to build and operate enterprise software AI-natively — a bet aimed squarely at the effort-based model the Indian IT-services majors sell.


SERVICES · FUNDING · STRATEGY · June 24, 2026

Vishal Sikka launches Hang Ten Systems with a $32M seed to undercut the IT-services model

Vishal Sikka, who ran Infosys as CEO from 2014 to 2017, came out of stealth on June 24 with Hang Ten Systems, a Palo Alto enterprise-AI-services startup, and a $32 million seed round led by Mayfield. Aramco Ventures joined as a strategic investor alongside participating angel investors; Yahoo co-founder Jerry Yang sits on the board. The company says it builds and operates enterprise software using AI-native methods — agentic code generation, reusable skills libraries, and domain-specific expertise — to cut the cost and time of building, customizing, integrating, and maintaining software. Siemens Gamesa Renewable Energy and Fresenius are named as early customers. Navin Budhiraja (CTO), Sanjay Rajagopalan (chief design officer), and Tao Liu (forward-deployed engineering) round out the founding team.

What this means. The pitch targets the exact cost structure the Indian IT-services majors sell. "Build, customize, integrate, maintain" is a four-line description of what TCS, Infosys, Wipro, and HCL bill for — mostly on effort, priced by people-hours, seat, and tenure. Hang Ten's claim is that AI-native build-and-operate collapses that effort. If it holds at scale, it compresses the thing the SIs charge for.

The credibility is specific. Sikka ran Infosys, knows the services P&L from the inside, and built and sold enterprise software at SAP before that. He is now selling against the model he ran.

The counts are early. Thirty-two million dollars is a seed round. Two customers are named. The category is crowded — Cognition's Devin and a wave of "the AI does the building" startups — and AI-native services has more announcements than audited outcomes so far. The thesis is plausible and the operator is credible; the evidence that this displaces incumbent engagement scopes does not exist yet.

India angle. IT and ITeS is among India's largest export sectors, and the SI majors are its anchor. A well-funded challenger built to undercut the effort-based build/integrate/maintain model bears directly on that franchise's structural competitiveness — which is why a US-domiciled, US-and-Aramco-funded startup is an India story at all. Hang Ten is not an Indian builder. The India exposure is competitive, sitting on the demand side of the SI book.

The incumbents are not static. Infosys, TCS, and Wipro pushed past 300,000 Microsoft 365 Copilot seats by early June, absorbing the same agentic-productivity vector Hang Ten is built around. LTIMindtree's BlueVerse Currency, introduced June 10, moves services pricing from effort to outcomes — the incumbent-side answer to the cost-collapse Hang Ten pitches. The contested question is not whether AI-native methods change services economics; the majors concede that. It is who captures the value when build/integrate/maintain costs fall — the new entrants, or the incumbents repricing their own books fast enough to keep it.

Behind the news. This sits on a thread running through 2026: capital and product converging on IT-services displacement. Cognition, maker of the Devin coding agent, raised $1 billion at a $26 billion valuation on May 27 — the canonical comparison for "autonomous coding compresses services revenue." Hang Ten sharpens that thesis from an unusual seat: a former chief executive of one of the largest Indian SIs, now selling against the model he ran. The novelty is the vantage, not the category.

What to watch. Whether Hang Ten names enterprise customers beyond Siemens Gamesa and Fresenius over the next few quarters — and whether any disclosed win is reported as displacing an incumbent SI's scope rather than greenfield work. That distinction, replacement versus net-new, is what separates another competitive services vendor from a threat to the export book. Watch also the Indian majors' FY27 pricing-model disclosures for how fast effort-to-outcome repricing spreads.

What this is not. Not evidence that the Indian IT-services model is being displaced. A $32 million seed and two named customers is a thesis with a credible author, not a demand shift. The frame that reads "ex-Infosys CEO will disrupt Indian IT" treats a funding event as a market outcome. The capability test is which engagements actually move, and none have been reported yet.

See also:

Source: TechCrunch, June 24, 2026; corroborated by Entrackr and Analytics India Magazine. → link

Confidence: High on the raise, investors, board, and stated product; the displacement implications are forward-looking.


Position movements

DimensionDirectionMagnitudeWhy
Enterprise adoption depth (SI services layer)01An AI-native build/operate challenger predicts pressure on the SI franchise's effort-based value capture, but a $32M seed has moved nothing yet — touched and predicted, direction unresolved.