Vishal Sikka's Hang Ten Takes Aim at IT Services With AI

Vishal Sikka Launches Hang Ten Systems on 24 June 2026

On 24 June 2026, Vishal Sikka, the former Chief Executive Officer of Infosys, publicly launched Hang Ten Systems, a Bay Area-based startup positioned as an AI-native alternative to the traditional IT services delivery model. The company simultaneously disclosed a 32-million-dollar seed round led by Mayfield Fund, with a strategic investment from Aramco Ventures. Jerry Yang, the co-founder of Yahoo, joined the board. Sikka led Infosys from 2014 to 2017 and subsequently founded enterprise AI company VianAI; Hang Ten Systems represents his most direct challenge yet to the incumbent model of how enterprise software is built, modified, and operated at scale.

Hang Ten began operating approximately one month before the public announcement and had already secured paying customers including Siemens Gamesa Renewable Energy and Fresenius within that first month. The speed of early enterprise traction indicates that large corporate clients are already prepared to trial AI-native services delivery — a posture that was not commercially credible even twelve months prior.

The Structural Problem With Traditional IT Services

The IT services market that Hang Ten is targeting is worth more than 200 billion dollars annually and built on a delivery model that has remained largely unchanged since the 1990s: clients pay services firms by the hour or by outcome for teams of engineers who write, maintain, and modify enterprise software. That model produced enormous businesses — including Infosys, TCS, Wipro, and HCL Technologies — but it contains a structural inefficiency that AI now makes visible and quantifiable.

Traditional IT services firms charge for the time it takes human engineers to understand an existing codebase, write new functionality, run tests, and maintain running systems. When agentic AI can autonomously generate code across multiple files, run its own tests, and propose fixes without human intervention at each step, the economic logic of billing for those human-hours erodes. The major services firms have recognised the risk and responded with AI partnerships — Infosys with Anthropic, Wipro with IBM — but their core businesses depend on large engineering headcounts in ways that make a fundamental structural shift difficult to execute from within established organisations.

Hang Ten Systems, starting without inherited headcount economics or existing client revenue to protect, is not constrained by those pressures.

How Hang Ten Systems Works

Hang Ten's delivery model is built around three components: agentic code generation, reusable AI skills, and domain expertise. Rather than deploying a team of engineers who spend weeks ramping on a client's existing codebase, Hang Ten uses AI agents to analyse codebases autonomously, generate new functionality across multiple files, and produce testable outputs without requiring human intervention at each step of the process. The reusable AI skills component means that capability built for one client engagement can be adapted for another, compressing the time-to-delivery curve on subsequent projects. Domain expertise — held by the founding team — provides the human judgement required for architecture decisions, stakeholder requirements, and the edge cases that agentic systems cannot yet handle without oversight.

Hang Ten's co-founders are Navin Budhiraja as chief technology officer, Sanjay Rajagopalan as chief design officer, and Tao Liu as senior vice president of forward deployed engineering. All three worked with Sikka previously, across SAP, Infosys, and VianAI, giving the team accumulated experience in building and deploying enterprise AI products at scale before this launch.

The India Connection

Hang Ten Systems is headquartered in California, not India, but its implications for India's IT services industry are direct and substantial. India's software services sector — anchored by Infosys, TCS, Wipro, HCL Technologies, and Tech Mahindra — employs more than five million people and generates the majority of the country's technology export revenue. That sector is built on exactly the human-labour-intensive delivery model that Hang Ten is designed to replace.

If agentic code generation can consistently deliver enterprise software outcomes that previously required large engineering teams, the headcount economics of Indian IT services firms face structural pressure that AI partnerships and internal retraining programmes will not fully address on their own. The risk that these companies have been managing as a medium-term concern is now arriving as present competition from new entrants built around the AI-native model from the start.

One publication headlined the launch as Sikka wanting to do what Infosys once paid him to prevent — a reference to the AI-driven automation of software delivery that Sikka attempted to accelerate during his Infosys tenure and that the organisation ultimately was not prepared to pursue at the pace he recommended.

What Indian Software Teams Should Watch

The speed of Hang Ten's early traction — paying clients including two large European multinationals within the first month of operation — is a signal worth taking seriously. It suggests that enterprise buyers have moved from evaluating AI-native delivery as an interesting future option to actively piloting it as a present-day alternative.

For Indian IT services firms of all sizes, Hang Ten should be read less as a single competitor and more as an indicator that a category of AI-native services firms is forming. The question for incumbents is not whether to respond but how quickly. For smaller Indian consultancies and product companies that are not constrained by the headcount economics of the large services firms, the Hang Ten model points toward a viable architecture: a smaller team of senior engineers directing agentic systems, pricing on outcomes rather than hours, and margins that do not depend on large bench utilisation rates.

For Indian engineering talent, the transition has two dimensions simultaneously: downward pressure on demand for traditional human-hours software delivery, and rising demand for engineers capable of directing agentic systems, evaluating AI-generated code at scale, and building the domain expertise layers that AI alone cannot yet provide.

The Bottom Line

Vishal Sikka, former Infosys CEO, launched Hang Ten Systems on 24 June 2026 with a 32-million-dollar seed round led by Mayfield Fund and strategic backing from Aramco Ventures. Co-founders Navin Budhiraja, Sanjay Rajagopalan, and Tao Liu bring prior experience across SAP, Infosys, and VianAI. The company's model replaces traditional human-intensive IT services delivery with agentic code generation, reusable AI skills, and human domain expertise, and had already secured Siemens Gamesa and Fresenius as customers within its first month of operation. For India's IT services industry, which employs more than five million people on the delivery model Hang Ten is targeting, the startup's launch and rapid early traction are a concrete signal that the AI-native disruption has moved from future risk to present competition.