Hang Ten, the four-month-old firm founded by former Infosys CEO Vishal Sikka, has raised $53 million in a second seed round led by Xora, taking total funding to $85 million. It sells AI-assisted software work to enterprises that typically clear $10 billion in annual revenue — Fresenius Kabi, Saudi Aramco and Siemens Energy are already on the client list — and says it staffs some projects with two to four people where roughly 30 specialists used to be needed. Sikka used to run one of the world's largest systems integrators. He is now selling the argument against that model to the companies that buy from it.
Mayfield, which led the first $32 million seed, came back for the second. The rest of the round is unusually operator-heavy: Aramco Ventures, Intel CEO Lip-Bu Tan, Micron CEO Sanjay Mehrotra, and Yahoo co-founder Jerry Yang, who sits on Hang Ten's board. Sikka would not give a valuation, saying only that the second seed priced above the first.
By his account, Xora came to Hang Ten rather than the other way around, prompted by early client results, and saw a chance to introduce the startup to other Temasek portfolio companies. That is worth reading plainly: part of what Hang Ten bought with this round is a distribution channel into a sovereign fund's holdings, with equity attached. The Aramco Ventures line and the Saudi Aramco client line in the same story point the same direction.
The commercial numbers are early but not trivial. Hang Ten has 21 large enterprises in play, some already holding proposals, some negotiating terms. It has several seven-figure contracts and projects in the tens of millions, spread across the US, Europe, the Middle East and Asia. In one case, 25 days passed between a first meeting and a multimillion-dollar contract for a critical software system — a speed Sikka says he has not previously seen from large corporate buyers.
The company is small for all this. Roughly 20 to 25 people work out of a Palo Alto headquarters plus the Middle East and Australia, with hiring planned in Europe and India. The new money goes to engineering, consulting and sales headcount.
Hang Ten is explicit that it is not training a model. Co-founder and chief design officer Sanjay Rajagopalan told TechCrunch that its engineers work through an in-house framework called Hobie, which bundles reusable AI skills aimed at regulated industries and complicated enterprise projects. The company says it ships software that runs in production rather than stopping at pilots. Sikka's own framing of the shift is that writing programs now costs almost nothing in incremental time and money, and that the work has moved from writing code to specifying requirements and verifying the system does what was intended.
That second half of the sentence is where the claim gets thin. Hang Ten's stated promise is a tenfold improvement in cost, speed, or some combination of the two — a target that can be met by any mix of the two and therefore cannot really be missed. And by the company's own description, final quality assurance and certification still sit with the client or an independent third party. In pharmaceuticals, energy and industrial equipment, verification is the expensive, slow, liability-bearing part of the job. Compressing the build from 30 people to four does not compress that, and nothing in the announcement says who is on the hook when a system written by four people fails an audit.
The displacement story is also softer than it looks. Sikka says many Hang Ten projects replace incumbent contractors, but more than half of current opportunities are new projects the client had previously shelved. That is a better business than it is a narrative: the firm is mostly selling work nobody was doing, not taking work off the incumbents' books. It also means the cost comparison — four people against thirty — is being made against a counterfactual, not a bill the client actually paid.
The most revealing fact in the story is the one with the fewest details attached. Four months after launch, with two dozen employees, Hang Ten received serious acquisition offers from what Sikka calls "very large companies," which he declined to name and declined to accept. Somebody with real money decided this was cheaper to buy than to answer.
Sikka has run this play before, with a different ending. He left Infosys in 2017, founded the enterprise AI startup VianAI in 2019, raised a $50 million seed and then $140 million in 2021 led by SoftBank Vision Fund 2. His LinkedIn profile shows he left VianAI in April; he told TechCrunch the company is "going through a transaction" and said nothing further. So the $190 million predecessor is being quietly disposed of in the same year the $85 million successor is being sold to overlapping buyers. Asked what firms like Hang Ten mean for service providers like Infosys, Sikka's answer was that his new company carries no legacy it has to transform.
The name is a surfing term — all ten toes on the nose of the board, which Sikka offers as an image of helping large companies stay upright on what he calls probably the biggest wave of this generation's lifetime. It is also, in the sport, the most exposed position you can take. Hang Ten's bet is that twenty-five people can underwrite the core software of ten-billion-dollar companies. The $85 million buys enough runway to find out whether contracts that took 25 days to sign hold up across the years those systems have to run.