Share with your CIO
Hang Ten Systems, founded by former Infosys CEO Vishal Sikka, is betting that large enterprises will pay a premium for an AI-native delivery model that combines agentic code generation with deep domain expertise, and $85 million in total funding suggests serious investors agree. The $53 million seed round was anchored by Temasek-backed Xora Innovation, with Mayfield and Aramco Ventures also participating. Hang Ten counts Aramco and Siemens Gamesa as customers and signed several multi-million-dollar contracts in the five weeks between its two funding rounds.
What this means for your business
The five-week contract sprint between rounds is the number that deserves attention. It signals that Hang Ten isn’t pre-revenue speculation; it’s a services firm already converting enterprise conversations into signed commitments at a pace that justified a follow-on check of this size. CIOs shopping for AI implementation partners should recognize the pattern: a credentialed founder with deep enterprise relationships (Sikka ran Infosys, a 300,000-person IT services giant) building a smaller, faster, AI-native version of the same business. That’s a direct challenge to incumbent SI relationships, and it’s worth knowing whether your current partners are building equivalent capability or just rebranding existing headcount.
Xora’s managing partner flagged the real constraint correctly: deployment speed, security, and economics are where enterprise AI adoption stalls, not model quality. Hang Ten’s reusable skills library, a pre-built collection of AI capabilities that can be assembled and adapted across engagements rather than rebuilt from scratch each time, is a direct answer to the economics problem. If it works at scale, it compresses the cost and timeline of AI delivery in ways that make the traditional time-and-materials SI model look expensive. The risk is that “reusable” is always the promise; whether the library actually generalizes across industries and stack configurations is what the next 18 months of customer results will prove or disprove.
Intel’s CEO and Micron’s CEO are personal investors here, which isn’t just a validation signal. It’s a tell about where Hang Ten expects compute intensity to land. AI-native software delivery at enterprise scale burns significant infrastructure, and having chip executives on the cap table suggests the company is already thinking about preferred access to silicon resources that most boutique AI firms can’t negotiate. CIOs renewing large SI contracts in the next budget cycle should weigh whether their incumbent partners have made equivalent bets on AI delivery architecture, or whether they’re defending margin on legacy delivery models while newer entrants accumulate the case studies.
Concept deep-dive: Agentic code generation
Agentic code generation means AI systems that don’t just autocomplete a developer’s prompt but autonomously plan, write, test, and iterate on software with minimal human intervention, closer to a junior engineer than a spell-checker. It exists because foundation models now have enough reasoning capability to handle multi-step programming tasks end to end. The business connection is direct: if agents can handle a meaningful share of software modification and maintenance work, the cost curve for enterprise IT delivery changes structurally, not incrementally.
Based on reporting from Temasek-backed Xora leads $53m funding for US-based Hang Ten, originally published 2026-09-16 20:19:00.
