Building AI-Native Services
We are entering the first era where services can scale without scaling headcount.
AI isn't just another technology wave — it's the first technology that can scale judgment and intelligence.
That changes everything for the largest part of the Australia and New Zealand economy: the A$2 trillion services sector.
Scaling Judgment
Every previous technology scaled production or coordination. AI scales judgment itself.
In services, the unit of production is a decision. A customs broker interpreting regulations and assessing risk. An accountant prioritising issues and escalating problems. A facilities manager determining which maintenance is urgent and which can wait.
This judgment work — prioritisation, exception handling, trade-offs, escalation, interpretation — is where all the value in services lives. It is also where all the cost lives.
For the first time, we can build systems that do more than execute predefined workflows. They understand context, apply judgment, handle exceptions, and improve with experience.
And crucially, AI does not just automate judgment. It compounds it.
Every decision made, every exception handled, every edge case resolved makes the system better. The 10,000th customs declaration is processed better than the first because the system has learned from the previous 9,999.
This is fundamentally different from human-based services, where judgment scales linearly by hiring more people. And it is different from SaaS, which standardised processes but could not capture or scale the judgment locked inside people’s heads.
The consequences are profound:
The quality-scale tradeoff disappears: AI captures expert judgment and deploys it everywhere at near-zero marginal cost.
The complexity-efficiency tradeoff disappears: AI thrives on complexity, getting better with more context and nuance.
The economics transform: labour-intensive businesses with 10–20% margins become technology-enabled platforms with 40–60% margins, while delivering better, faster, more sophisticated services than were previously possible.
The Model Gap
The opportunity is clear. But almost no one is positioned to capture it:
Traditional service businesses have customers, operations, and deep domain expertise, but lack the product vision and technical capability to reimagine themselves. They struggle to attract world-class AI talent and rarely have the capital or patience for multi-year transformation. As a result, they try to “add AI” to existing workflows instead of fundamentally redesigning their business models.
Venture-backed software startups usually have the talent to build great tools. But they lack access to proprietary data and embedded business context, the inputs required to build truly fit-for-purpose AI systems. And because they sit outside the organisation, they do not have the authority to re-architect operations or drive fundamental change.
Private equity firms can acquire and optimise, but they're structured around financial engineering, not technology-led transformation. They also struggle to attract world-class AI talent and their 3–5 year hold periods create misalignment with the patient, compounding value creation that AI transformation enables.
Consultants promise transformation but don't own the outcome. They can't build and maintain AI systems because they don't operate the businesses. They can't capture compounding benefits because they leave after the engagement.
What's missing isn't technology or capital. It's a model that can execute transformation at scale — and capture the compounding value created.
A New Category of Firm
The most powerful model for building category leaders pairs operating ownership with embedded intelligence: AI-native services platforms.
Not services businesses experimenting with AI. Not AI vendors selling into services. A new category defined by:
Active Ownership. You can't build systems that scale judgment from the outside. You need proprietary access to how decisions actually get made — the context, the edge cases, the trade-offs. This level of access only comes from owning the business.
Vertical Intelligence. Unlike horizontal SaaS, you go deep in specific industries — building domain-specific intelligence that becomes a compounding moat.
Patient Capital. AI-native transformation creates value that compounds over decades, not quarters. You need capital aligned with building enduring platforms, not flipping assets.
Full-stack Execution. World-class product and engineering teams building sophisticated AI systems, combined with operational expertise from people who understand how to run and transform service businesses. This combination rarely exists.
This model also attracts the best professionals in each industry, because it can offer both the highest compensation and the most powerful infrastructure. AI absorbs the routine work, freeing top performers to focus on high-impact judgment and relationship-driven work, with compensation to match their impact.
Unlocking Potential at Scale
When judgment becomes software, the economics of services fundamentally change. Value compounds across multiple dimensions at once:
Revenue expansion. Better service delivery unlocks customers and offerings that were previously uneconomic, expanding addressable markets and improving acquisition and retention.
Margin transformation. At scale, AI absorbs judgment-heavy work that once required headcount, driving 20%+ EBITDA uplift through structurally lower operating costs.
Multiple expansion.You acquire at services multiples and build toward technology multiples as unit economics shift from labour-intensive to software-like.
These advantages compound further through a flywheel across verticals. The second accounting firm is faster to transform than the first, because the hard problems have already been solved.
Patterns learned in accounting inform how you approach customs brokerage. The meta-capabilities, mapping decision flows, extracting domain knowledge, building intelligence layers, become reusable infrastructure across industries.
That is how you build a multi-billion-dollar platform.
Why Services, Why Now
Services represent the largest part of the economy, roughly fifty times larger than software, spanning industries like accounting, logistics, and legal.
Yet services have seen little structural transformation in decades. The reason is simple.
Service work is messy, context-dependent, and judgment-heavy. Workflow tools improved coordination, but they could not handle the complexity that drives real delivery.
Software could organise work. It could not do the work.
Until now.
Introducing Dragonfly
Dragonfly acquires established service businesses across Australia and New Zealand and transforms them into category-defining platforms by embedding AI directly into core operations.
We’re building a new category of firm, one that owns operations, embeds intelligence, and compounds judgment across industries over decades.
This model combines operating ownership with world-class AI product development, patient long-term capital, and deep vertical expertise.
For decades, services and technology have been separate categories. That distinction is ending. The firms that own this transition will become the platforms of the next era.
