Enterprises across APAC are moving beyond GenAI experimentation to focus on data, infrastructure, and clear use cases that generate measurable business value.
As digital transformation accelerates across Asia–Pacific, Generative AI (GenAI) is shifting rapidly from technological fascination to a concrete business imperative. At the Canalys Forum APAC 2025, this transition took center stage during the Expert Hub session “The GenAI Advantage: From Pilot to Profit,” moderated by Sheena Wee, Principal Analyst at Omdia. The discussion brought together leaders from Novare Technologies Inc., VLAN Asia, Vertiv, Tech Data APAC, IBM, and Microsoft, offering a comprehensive look at how organizations can convert GenAI proofs-of-concept into measurable financial gains.
Experts agreed that GenAI creates value only when enterprises invest effectively in data foundations, scalable infrastructure, and clearly defined use cases—rather than relying on impressive demos or trend-driven experimentation.
A Shift From Curiosity to Core Business Needs
According to William Emmanuel Yu, CTO of Novare Technologies, most enterprise conversations around GenAI often begin with curiosity about tools like ChatGPT but quickly pivot to deeper issues: data governance, data lake architecture, and data quality. He emphasized that GenAI’s true potential emerges only when organizations identify well-scoped use cases and eliminate fragmented or unstructured datasets that hinder meaningful outcomes.
Execution Speed Determines Value Delivery
From an implementation standpoint, Lance Cheang, Managing Director of VLAN Asia, emphasized speed of execution as a decisive differentiator. Customer service, he noted, remains one of the clearest domains where GenAI can generate immediate impact—through rapid issue detection, classification, and intelligent response suggestions. However, he cautioned that human oversight remains essential, as no AI system can replace human judgment in quality assurance.
The Overlooked Challenge: AI-Ready Infrastructure
Representing the infrastructure perspective, Daniel Sim of Vertiv pointed out a gap many enterprises overlook: the physical backbone required to support GenAI workloads. As compute density rises from 10 kW to 50–100 kW per rack, traditional data centers are reaching their limits. Organizations must now rethink cooling, power distribution, and rack design to build truly “AI-ready” facilities capable of supporting large-scale deployments.
Starting With Clear Goals — Not Hype
From the distribution ecosystem, Mohamed Noorul Huq of Tech Data APAC warned against implementing GenAI out of competitive pressure or novelty. He urged businesses to begin with precise objectives and leverage ready-made solution packages to minimize cost and time to market. With more than 50 validated use cases available, distributors can act as strategic guides to ensure enterprises start in the right direction.
IBM: Reinvent Processes, Don’t Patch Them
Purushothama Shenoy of IBM highlighted that GenAI only delivers real value when it redefines core processes rather than simply augmenting outdated workflows. Enterprises need to use GenAI to drive agility, reduce cycle times, and modernize the way work gets done. He also underscored the importance of trust and transparency, which remain critical for customer confidence in AI-driven operations.
Microsoft: Services Sector Will See ROI First
Closing the session, Gerald Leo, ASEAN Channel Sales Director at Microsoft, noted that the services sector is best positioned to see rapid returns from GenAI. Automating repetitive tasks frees human talent for high-value work, boosts operational speed, and reduces errors. He recommended that businesses start small, build internal expertise, scale gradually, and collaborate with partners to optimize costs and resources.
APAC Enterprises Are Optimistic — But Barriers Remain
Recent surveys reflect this regional momentum: 91% of ASEAN enterprises believe GenAI will significantly reshape their operations within the next 18 months. About one-third have already deployed GenAI in production environments, while the majority remain in testing phases. However, compute-intensive workloads, volatile cloud expenses, infrastructure gaps, and talent shortages continue to restrain GenAI from achieving full potential. These challenges are prompting more organizations to consider AI-dedicated data centers to reduce dependency on public cloud platforms.
A key data point presented at the Forum showed that 87% of enterprises expect their AI budgets to increase in 2025, signaling strong confidence in AI-driven business models and automated operations.
A Roadmap for Vietnam’s AI Ambitions
For Vietnam—where demand for data centers, digital infrastructure, and AI applications is rising—the “pilot-to-profit” lessons shared at Canalys hold significant relevance. Vietnamese enterprises must prioritize:
- Strong data foundations
- AI-ready infrastructure
- Well-chosen use cases
- Sustainable talent development
This approach will ensure GenAI evolves not merely as a tool for experimentation, but as a long-term engine for growth and competitive advantage.