The conversation around AI is changing from tools and models to scale, deployment, and real-world impact. This is reflected in capital flows as 2025 becomes the third-highest year for global VC funding, with AI-driven startups in SEA securing over 30% of funding, totalling more than $2.3 billion. This signals a clear investor preference for startups that enable AI to scale beyond experimentation into sustained deployment.
Investors are increasingly backing AI solutions designed for infrastructure-level integration rather than incremental automation gains. We’re most excited about companies that aren’t using AI as a feature. They’re building with it as a core architectural assumption.
The blueprint for a new economic architecture
What makes these companies compelling is the structural advantages they create and the value they unlock. With AI projected to contribute 13–18% uplift to Southeast Asia’s GDP by 2030 (~$1T impact), it’s becoming a key driver of new demand across industries. As AI expands the market, the physical and digital infrastructure (compute, energy, data, and networks) becomes the foundation that enables it to be viable and scalable.
Serving as a core economic layer, an AI-underpinned economic architecture gives investors access to ecosystem-wide growth, not just single-use outcomes, offering more stable and diversified returns. Through shared infrastructure layers, multiple industries can build on top of the same systems, making it more valuable as more players depend on it. This scale extends across fragmented markets in SEA, enabling faster cross-border deployment and monetisation, positioning AI as the next structural layer of growth. In a market increasingly defined by fundamentals and clear paths to profitability, investors are backing companies that can operationalise AI into defensible businesses.
Investors who recognise that infrastructure is evolving into a critical economic layer should prioritise exposure to companies building these layers early. The opportunity lies in backing founders before valuations fully reflect regional and global potential.
Infrastructure in action: shifting the AI stack
Take Featherless AI, a serverless inference platform designed to host and run open-source AI models. By removing infrastructure complexity, the platform helps lay the foundation for accessible AI, allowing teams to focus entirely on independently building the software they need.
On the other hand, ORCA by Lydia AI, is a solution-intelligence platform that functions as an operational backbone for complex enterprise environments. It translates fragmented information and domain expertise into scalable, concrete solutions across industries. These examples signal how the AI stack is evolving toward infrastructure-led scalability, concentrating advantage in the systems that underpin entire workflows.
Enterprise adoption and execution is the true differentiator
Although early momentum is important, lasting value depends on whether a startup can withstand enterprise adoption: a key signal for both investors and LPs evaluating long-term potential. For investors, it underscores the importance of finding founders with the technical depth to translate AI capabilities into commercially viable solutions. At the same time, enterprises are becoming more selective as they move beyond AI-assisted processes toward AI-orchestrated execution that can adapt, learn, and run enterprise processes in real time.
As the bar for enterprise adoption rises, those who can identify and invest early in these solutions are best positioned to navigate these shifts and translate AI adoption into sustained competitive advantage. What will ultimately differentiate outcomes across the ecosystem is whether they treat AI infrastructure as a driver of growth rather than a cost constraint. Recognising these shifts early matters, especially while the playbooks are still being defined. As this unfolds, the next phase of AI might no longer be dominated by access but by execution and the ability to support it within real systems at scale.
The article titled “Why AI infrastructure is becoming Southeast Asia’s next competitive advantage” was authored by Mike Maté, General Partner at Kickstart Ventures
About the author
Mike is a General Partner at Kickstart Ventures, the Philippines’ largest corporate venture capital firm.
In this role, he has led significant investments, including a multimillion-dollar investment into UK-based Roslin Technologies, which develops animal cells for the cultivated meat sector, and Montreal-based SRTX, which has developed the world’s only full-stack, closed-loop, apparel manufacturing platform. He also sits on the board of Roslin as an observer, and Lydia AI, a Toronto-founded leader in health and actuarial sciences AI for real-time life and health underwriting, whose pre-Series B round Kickstart led.
Before joining Kickstart Ventures, Mike worked as a Director at Seawood Resources, a multi-billion peso investment firm, where he formulated investment strategies and oversaw the firm’s portfolio companies, including sitting as director on the board of Akaroa Salmon, a premium King Salmon producer in New Zealand.
Prior to his roles in the investment industry, Mike gained valuable experience in finance and law. He served as the Head of Structured Finance and Debt Products at BPI Capital, the investment banking group of Bank of the Philippine Islands. He also worked as a Senior Associate at Romulo Mabanta Buenaventura Sayoc & De Los Angeles, a prominent law firm in the Philippines. Additionally, Mike was a Visiting International Attorney in the M&A Group at Morrison Foerster, an international law firm with offices in the Americas, Europe and Asia.
Mike holds a Master of Laws (LL.M.) degree from Harvard Law School, for which he was the LL.M. Class Representative during his time there. He also earned a Bachelor of Arts (B.A.) degree in History from Duke University and a Juris Doctor (J.D.) degree from Ateneo de Manila University.