On June 24, 2026, Vishal Sikka — former CEO of Infosys, former CTO of SAP — launched Hang Ten Systems with a $32 million seed round to deliver what they call "AI-native project delivery" to large enterprises like Siemens and Fresenius.
The venture capital thesis is straightforward: AI has fundamentally broken the economics of building enterprise software. What used to require months of configuration, customisation, and integration can now be delivered in days. Bespoke software is no longer a luxury reserved for companies with IT budgets in the millions.
That thesis is correct. And it applies just as directly to businesses that are not Siemens.
What "AI-Native Delivery" Actually Means
The traditional model of enterprise software looks like this: a vendor builds a generic product, the customer spends months (sometimes years) configuring it to approximate their actual workflow, consultants handle the gaps, and the business eventually settles for a process shaped around the software rather than the other way around.
Hang Ten's pitch — and the broader industry shift it reflects — is that this model is obsolete. Agentic AI can generate, modify, and maintain code at a fraction of the previous cost and time. The marginal cost of building something that fits your specific workflow has dropped dramatically. "Bespoke" no longer means "expensive."
In Sikka's own words: "A few are already reaping massive benefits, building in days what used to take years. But most are stuck at the starting line, or worse, and the gap is widening every day."
He was talking about Fortune 500 companies. But the dynamic he describes — the gap between businesses that are using AI to real advantage and those still running on spreadsheets and workarounds — is even more pronounced for small and mid-market businesses.
The Same Economics, Without the Enterprise Price Tag
Hang Ten's model relies on three components: agentic code generation, a reusable skills library of domain-specific AI modules, and specialist delivery teams who supervise and apply them. The combination lets them deliver sophisticated enterprise capability faster and cheaper than before.
The same three components exist at smaller scale — and at lower cost — for growing businesses:
Pre-built industry modules. The equivalent of a "reusable skills library" for an SMB is software that starts from the right industry model. A boutique garment enterprise doesn't need a generic ERP adapted to fit their production workflow. They need software that already understands how orders, manufacturing logs, and revenue analytics fit together in that specific industry. That specificity is the starting point, not the end result of years of customisation.
AI built into the workflow, not wrapped around it. The AI capabilities in genuinely AI-native products — reporting automation, operational chatbots, document processing — are part of the system architecture, not features added to a feature list. When a business asks "where is my order?" and an AI assistant can answer by querying the live production database, that is not a chatbot feature. That is the system working the way it should.
Delivery that ships, then iterates. The AI-native model is not a big-bang implementation. It is a continuous delivery model: ship the core, measure adoption, iterate. That model works better for growing businesses than the traditional "implement everything, go live in six months" approach that suits large IT programmes but routinely fails at SMB scale.
Why the Gap Matters More at SMB Scale
Hang Ten's investors note that within a few years, the gap between enterprises that use AI to real advantage and those that don't will define entire industries.
That gap is already visible at SMB scale — and the consequences arrive faster. A small business that is still running operations on spreadsheets and WhatsApp is not just operating inefficiently. It is making decisions with incomplete data, losing time to manual work that could be automated, and falling further behind competitors who have better systems.
The businesses that automate intelligently now are building operational foundations that compound. Better data leads to better decisions. Better decisions lead to faster growth. Faster growth creates more data and more capacity to automate further.
The businesses that wait are not standing still. They are falling behind on a trajectory that accelerates with time.
What This Looks Like in Practice for a Growing Business
The practical version of AI-native delivery for a small or mid-market business is not an enterprise transformation programme. It is a single, focused starting point:
- A garment manufacturer deploys production management software with an AI assistant that answers operational questions — daily output, job status, efficiency by operator — without anyone running a report manually.
- A field service business deploys scheduling and job management software where dispatchers can ask the system for available engineers, check SLA windows, and generate client summaries in natural language.
- A textile producer connects production records, yarn movement, and machine tracking into a unified system where reporting that used to take hours happens automatically at the end of each shift.
None of these are enterprise transformation projects. Each one is a specific business problem with a specific software solution that includes AI where it removes real manual work.
That is AI-native delivery at SMB scale. Not a concept. Not a roadmap item. A deployable system that runs in production.
The Honest Caveat
Not every AI capability is worth deploying. The phrase "AI-native" is already attracting the same hype that "digital transformation" attracted a decade ago — and the same risk of initiatives that consume budget and produce dashboards instead of outcomes.
The right test is simple: does this remove real manual work that a human is currently doing? If yes, it is worth building. If it produces insights that require another human to interpret and act on before anything changes in the business, the ROI will be thin.
Hang Ten is targeting the AI capabilities that change how enterprise software is built and maintained. The same discipline — AI where it removes genuine friction, not AI as a feature — applies to every business that is evaluating what to automate next.
For a practical breakdown of how we implement AI-native delivery — including specific workflow automations, the Claude API integrations we build, and the delivery model we use — see our AI Automation & AI-Native Delivery service page.
Ibistra Tech builds AI-native software for small and mid-market businesses in garment manufacturing, textile production, field service, and SaaS. Talk to us about your specific workflow.