As data centres become the core of the AI economy, their greatest constraint is no longer compute, it’s power. For investors, founders, and operators in Southeast Asia, this convergence of artificial intelligence and energy presents both a bottleneck and a generational opportunity. This is our view from The Radical Fund: the next frontier of digital progress will come from climate-aligned infrastructure that fuses intelligence with power.

Over the following sections, we explore:

• The challenge: How exponential data and compute demand are colliding with the physical limits of energy systems
• The shift: Why efficiency alone is no longer enough, and how system-level innovation is reshaping the data centre model
• The regional lens: Why Southeast Asia, with billions in FDI and a fast-digitising economy, is uniquely positioned to lead this transition
• The opportunity: Where investors and founders can enable scalable solutions in cooling, compute, energy integration, and grid resilience

Digital abundance meets physical constraint

Artificial intelligence is redefining the boundaries of computation, data, and productivity. Yet behind every model, every query, and every algorithm sits a data centre consuming colossal amounts of power. Global data centre demand is projected to more than double by 2030, rivaling Japan’s total annual electricity use. The infrastructure built to enable the digital economy is colliding with the physical limits of energy systems that were designed for a different century.

Energy has become the new bottleneck of AI progress, the physical constraint in a digital race. Hyperscale facilities that power today’s cloud and AI workloads already account for roughly 1.5 per cent of global electricity consumption. This share is growing quickly as AI training and inference workloads multiply.

Nowhere is this constraint more visible than in Southeast Asia. The region is witnessing a surge of digital activity and foreign direct investment into hyperscale data centres. Malaysia, Indonesia, and the Philippines are positioning themselves as new digital gateways. Malaysia alone has announced more than MYR 99 billion, or US$23 billion, in planned data centre investments since 2023. Indonesia and the Philippines are following closely behind. Yet the regional grid remains fossil-heavy, underinvested, and unevenly modernised.

Also Read: Energy business, the engine of sustainable global transition

The world’s most advanced computation networks are running on infrastructure built for another era. Without rapid innovation at the intersection of energy and intelligence, the very systems driving the AI revolution could face their own energy ceiling.

The energy makeup of intelligence

A modern data centre is, in essence, a miniature energy ecosystem, with roughly 40 per cent of total energy use going to compute, and another 40 per cent to cooling. Both are rising sharply as high-performance GPUs replace traditional CPUs and as AI workloads scale.

Each hyperscale facility now draws as much power as a small city. Johor, Malaysia’s emerging AI hub, could account for nearly 30 per cent of national power consumption by 2030 if all planned capacity comes online. The concentration of demand is staggering.

Power Usage Effectiveness, or PUE, has long been the industry’s benchmark for efficiency. A perfect score of 1.0 means every watt powers computation alone. Yet even the most advanced facilities, with PUE ratios near 1.1, face a bigger challenge: total power demand is compounding at double-digit rates. Incremental improvements can no longer offset exponential growth. The conversation must shift from energy saving to system redesign.

This is not just a sustainability issue. It is an economic one. Energy costs account for between 30 per cent – 50 per cent of total data centre operating expenses. As power tariffs rise and emissions rules tighten, energy strategy becomes synonymous with business strategy.

The required system-level shift from efficiency to integration

The data centre industry has long approached sustainability as a collection of independent problems: efficiency on one side, compute on another, and grid supply somewhere outside the fence. That era is ending. The next generation of digital infrastructure will be designed as an integrated system, where power, heat, and compute flow dynamically across the same operational stack.

Southeast Asia offers fertile ground for this transformation. In Singapore and Malaysia, operators are testing liquid and immersion cooling systems capable of handling the extreme thermal densities of AI chips. These technologies replace traditional air-conditioning with precision systems that use water or non-conductive liquids to extract heat directly from processors. In a region where temperatures are high and land is scarce, the shift from air to liquid cooling can reduce cooling energy use by roughly a third, according to industry benchmarks, while freeing up space for more compute.

Integration extends beyond cooling. Graywater recycling and waste-heat recovery are becoming viable in data parks connected to urban industrial clusters. In Singapore, treated wastewater already accounts for over 40 per cent of the national supply, setting a foundation model for closed-loop cooling systems. In cooler regions such as Europe, wasted data centre heat is being reused in district heating systems. In time, Southeast Asia may find its own circular approaches suited to tropical climates and water scarcity.

The most significant leap will come from software addressing incompatible systems. Digital twins and real-time analytics platforms are emerging to orchestrate infrastructure dynamically, predicting load shifts, adjusting cooling, and optimising power flows without new hardware. This software-defined approach blurs the line between IT and energy operations, creating adaptive, self-optimising systems. Efficiency becomes not a fixed objective but a continuous function.

Energy independence as a strategy

Even as integration advances, the grid itself is becoming a constraint. Across Asia, grid connection delays now exceed data centre build times. In hotspots like Johor and Batam, connection queues stretch for years. Meanwhile, fossil price volatility, emission caps, and renewable intermittency have made energy planning both more complex and more strategic.

Forward-looking operators are responding with on-site generation and storage, together with hybrid power models that provide autonomy and resilience:

• Co-located solar and battery systems that offset daytime load and stabilise supply;
• Hydrogen-ready microgrids that future-proof against fossil fuel volatility; and
• Small modular reactors (SMRs) are being explored for stable, round-the-clock baseload power.

These models reduce exposure to fossil volatility and regulatory tightening, providing the ability to stay online when the grid cannot.

Energy independence is fast becoming a driver of valuation. Facilities that adhere to renewable integration standards, interconnection requirements, and carbon-reduction mandates face lower operational risk and, therefore, lower weighted average cost of capital. For investors, this translates into higher exit multiples. What began as environmental compliance is now a form of financial resilience.

The narrative is evolving from green compliance to energy resilience, from sustainability as an obligation to sustainability as a competitive advantage. The AI revolution will not be won in the cloud, but in the power grid that sustains it.

Also Read: The shifting geopolitics of sustainability, energy, and climate

The new asset class: AI-ready infrastructure

A new category of infrastructure is emerging, one that is intelligent, efficient, and sovereign. Energy-smart data centres will define the 2030s, and the convergence of compute, energy, and regulation will shape not only the digital economy but also national competitiveness.

Southeast Asia is already becoming a stage for this transformation. Singapore remains the premium ESG benchmark, but with limited land and water, it is guiding regional standards rather than expanding capacity. Malaysia has seized the opportunity, attracting a wave of investment from global hyperscalers. Indonesia is rising fast, driven by its massive population and government incentives. The Philippines and Vietnam are catching up as connectivity improves.

This FDI surge is more than a capital inflow. It signals a strategic repositioning. Nations are competing not just to host data but to control the digital-physical nexus of energy and computation. The outcome will determine who captures the value created by the AI economy.

Investors, policymakers, and builders are no longer operating in silos. They are co-designing an ecosystem where energy efficiency, grid intelligence, and data sovereignty intersect. For capital allocators, this presents a generational opportunity: to fund the foundations of an AI-ready, climate-aligned digital economy.

The next decade

The next decade will test whether the world can reconcile digital expansion with environmental limits. The AI era is not merely a software story; it is an energy story. Without reinvention at the infrastructure level, capacity, cost, and carbon will become binding constraints on innovation.

Southeast Asia stands at the forefront of this challenge. Its economies are growing rapidly, its populations are digital-first, and its geography places it at the crossroads of East and West. Yet its energy systems remain among the most carbon-intensive. Bridging that gap requires imagination and investment in equal measure.

This region can lead by designing the next generation of infrastructure from first principles, embedding energy intelligence into every layer of the digital stack. Governments can align data-centre policy with national energy transition plans, accelerating renewable integration and storage. Investors can support technologies that couple compute density with sustainability. Operators can adopt circular resource models for heat, water, and hardware.

Southeast Asia has the resources, capital, and talent to shape this future. The question is whether it will choose to lead or wait for others to define the standards.

At The Radical Fund, our view is clear. The AI revolution depends on reinventing energy infrastructure. The region that succeeds in aligning power with intelligence will not only fuel its digital growth, but it will also own the foundations of the next economy.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. Share your opinion by submitting an article, video, podcast, or infographic.

Join us on Instagram, Facebook, X, and LinkedIn to stay connected.

Image credit: Canva

The post Why the AI revolution depends on reinventing energy infrastructure appeared first on e27.

In my decade and a half of marketing experience, I’ve witnessed how the explosion of digital tools and data has revolutionised marketing. Yet, despite all these advancements, many businesses—including those in retail, F&B, and technology—still struggle with a fundamental challenge: understanding what truly drives consumer decisions.

Halfway into 2025, most brands have yet to recognise the power of AI, beyond ChatGPT; AI is the essential driver for turning data into real marketing success.

From data overload to actionable insights

Every day, companies collect mountains of data—clicks, views, social mentions, and more. But too often, the story ends there. Many brands get stuck chasing vanity metrics, missing the deeper insights that reveal why customers act the way they do. In my experience, this is where most businesses falter: they have the data, but not the understanding.

Consider this: 84 per cent of companies in Singapore use digital channels to promote their products and services, but only 17 per cent can directly link their marketing strategies to increased revenue.

That’s far below Asia’s regional average of 41 per cent. For me, this highlights a critical gap. It’s not about having more data; it’s about making sense of it in a way that drives real, measurable results.

Why traditional marketing methods aren’t enough

Modern marketing isn’t about following the latest trends or relying on gut instinct. Success comes from making informed decisions based on a deep understanding of consumer behaviour. When businesses stick to outdated methods or surface-level analytics, they risk missing what truly drives conversions, loyalty, and growth.

How AI and data analytics are changing the game

AI and advanced analytics have become game-changers for brands. Take Coca-Cola, for example—they use AI to analyse customer preferences, buying habits, and social media trends to create hyper-targeted campaigns and new products. Amazon goes even further, applying AI for personalised recommendations, dynamic pricing, and real-time inventory forecasting. These strategies have boosted both customer satisfaction and sales.

Also Read: VC crunch hits Southeast Asia: US$129M raised in May 2025, down 70% MoM

Unlike traditional market research, which I’ve seen can be slow and expensive, AI lets us interpret vast amounts of complex data in real time. This means we can adapt marketing strategies based on what consumers are actually doing—not just what we think they might do.

What to expect from data-driven marketing solutions

As competition grows fiercer, I know that brands—my own included—expect more from their marketing investments. We need data platforms that analyse consumer behaviour across multiple touchpoints and deliver actionable insights. It’s no longer enough to track page views or clicks. I want to understand the “why” behind customer actions.

Understanding consumer psychology and predicting behaviour has become a personal focus for me as a marketer. With shifting preferences, fragmented channels, and rising expectations, I can’t afford to rely on surface-level metrics. I need to know what truly motivates my audience to engage, convert, and stay loyal.

Adopting a behaviourally informed approach is non-negotiable. We help brands go beyond performance metrics to tap into deeper insights—how customers think, what influences their decisions, and how to communicate with them more meaningfully.

Making data-driven marketing a strategic asset

For me, embracing AI-driven insights isn’t just an option—it’s a necessity. McKinsey and KPMG both highlight how data-driven marketing is about making smarter, more effective decisions that align with consumer needs. The brands I see succeeding are those leveraging these insights to improve retention, boost engagement, and use their marketing budgets more efficiently.

Future belongs to smart solutions

If you want to stay ahead of the competition and drive real, sustainable growth, investing in AI-powered, data-driven marketing is the way forward. With the right insights, you can transform your marketing strategies and build stronger, more lasting connections with your customers.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. Share your opinion by submitting an article, video, podcast, or infographic.

Join us on Instagram, Facebook, X, LinkedIn, and our WA community to stay connected.

Image credit: Canva Pro

The post Why AI is essential to understanding consumer behaviour for marketing success in 2025 appeared first on e27.

Seeking a state of calm in the current trade storm is clearly challenging the minds of corporate strategists around the globe today, but one thing remaining constant is the focus and optimism over AI and digital capabilities.

Amid uncertainty, digital transformation remains a critical driver for business and deal strategies, with AI capabilities increasingly driving corporate acquisitions. At the same time, a shift to defensive consolidation helps companies build operational and competitive resilience.

The EY-Parthenon CEO Outlook Survey revealed unprecedented optimism in the transformative power of technology, particularly AI, from Asia-Pacific’s CEOs. In fact, 85 per cent predict that AI will be a decisive factor in establishing industry leadership in 2025. An even higher number, 87 per cent of leaders, believe that the accelerated adoption of AI and the associated up-skilling of their workforce will be crucial differentiators in the coming years.

The latest EY research on GenAI shows 90 per cent of firms surveyed have adopted AI into operations to some degree, but most remain in the early stages, and eight per cent have not integrated AI at all.  72 per cent of surveyed leaders plan to increase annual investment related to GenAI specifically, and the rush to AI challenges business leaders to seek both optimisation of adoption strategies and acquisition as a means to accelerate capabilities.

M&A outlook in flux but opportunities remain

The prevailing optimism around AI and technology coincided with an early surge in expectations in M&A deals at the start of the year. Nearly half of the CEOs surveyed (42 per cent) expressed strong confidence in investing in Capex and R&D, as they focus on leveraging technological advancements such as AI to secure competitive advantages.

Many CEOs identified M&A as a transformation accelerant in 2025, with confidence by Asia-Pacific CEOs (61 per cent) driving an even greater appetite for deals versus European and American peers.

Also Read: When the chain snaps: How tariffs are unraveling Southeast Asia’s SMEs

In response to significant trade and market disruption, a portfolio realignment is essential to counter geopolitical risk and shifting market growth potential. Trade policy and supply chain risks will now govern many investment decisions.

 CEOs need to navigate this landscape adeptly to fully harness the value of any planned transactions while integrating technology investments that reshape growth strategies.

Pockets of opportunity remain, such as in Southeast Asia. Despite risks from recent tariff announcements out of Washington, the region could see increased investment and partnerships as governments and businesses seek growth opportunities in the volatile landscape.

Businesses anticipate greater investments through joint ventures, partial ownership, and minority interests in Asian companies. This reflects their need to balance uncertainty in countries like China, Canada, Mexico, and the EU that face added complexity given implications of tariff announcements and growing trade challenges. Strategic partnerships will help businesses navigate risks and drive growth.

There are clear opportunities in markets and regions of growing domestic demand to “build” domestic supply chain ecosystems to avoid tariff risks. The major obstacle in this path is that this strategy hinges on more than procuring and manufacturing locally, but also whether the destination of finished goods can be locally contained.

Also Read: Asia’s trade turning point: How tariffs and geopolitics are redrawing supply chains

More mature markets, like Japan and Korea, face the complex challenge of consolidating and restructuring their companies’ supply chains, which have expanded globally both upstream (e.g., raw materials and production) and downstream (e.g., distribution and sales). In a tough economic climate, achieving cost efficiencies through these efforts becomes increasingly attractive.

Seeking a state of hyper-agility and resilience

Companies will view current critical trends and disruptions as either threats to their existence or opportunities. It is this difference that separates industry leaders from laggards. Today’s business landscape is shaped by disruptive forces: rapid technological advancements, including artificial intelligence (AI), climate change-driven sustainability agendas, and geopolitical tensions affecting supply chains and global operations.

Digital ecosystems expose new threats and opportunities, while remote work reshapes organisations. Rising cybersecurity risks, shifting consumer expectations, economic volatility and complex regulations demand agility. Meanwhile, emerging markets create new competitive dynamics and growth potential.

In terms of deal strategy, while uncertainty clouds the immediate future, M&A should still be leveraged as a transformation catalyst. CEOs should look to M&A, particularly in distressed times as an opportunity for accelerated transformation. Target deals will include those that align with long-term goals, such as adopting new technologies, entering new markets or strengthening competitive positioning through strategic consolidation.

CEOs clearly need to sharpen their focus on the interplay of macroeconomic, geopolitical, regulatory and technological forces. Proactively addressing these risks enables growth opportunities and helps mitigate disruption. The traditional business dashboard must factor in more diverse sources of insight and more outlier views than in the past. New indicators must be established to build a wider view of possible scenarios so that a path to operational resilience can be found.

The views reflected in this article are the views of the author and do not necessarily reflect the views of the global EY organisation or its member firms.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. Share your opinion by submitting an article, video, podcast, or infographic.

Join us on Instagram, Facebook, X, LinkedIn, and our WA community to stay connected.

Image credit: Canva Pro

The post From uncertainty to action: Power of AI and digital shaping deal strategies in turbulent times appeared first on e27.

As AI adoption accelerates across Asia Pacific, the region is facing an urgent need to implement more sustainable AI practices. AI workloads – particularly those running on hyper-scale data centres – are energy-intensive. Given that Asia Pacific has accounted for the largest share of newly added data centre supply over the past decade, ensuring AI systems are sustainable is a growing priority.

AI’s environmental impact is a subject that is well-established, no longer a conversation limited to expert forums and panel discussions. Although AI technologies promise breakthroughs in healthcare, climate science, energy transition acceleration, and more, it’s crucial to address AI’s predicted environmental downsides.

At its core, AI models are trained and run on extremely powerful, energy-hungry computers that rely on electricity that largely comes from carbon-intensive sources. Left unchecked, this could lead to significant carbon emissions.

There are many ways that we can address AI sustainability, but one often overlooked lever we can pull to improve the efficiency of AI workloads is data optimisation, or “data efficiency”. AI models rely on vast pools of data to be effective, but indiscriminately dumping disorganised, irrelevant, or even duplicate data into AI models leads to systems having to do extra work processing excess information. We cannot afford to be wasteful when it comes to AI.

By optimising the data before we feed it to AI models, we can help better manage the environmental footprint of AI. This requires careful forethought and expert planning that looks for sustainability gains along the entire AI lifecycle and prioritises data efficiency when planning AI projects.

How to tackle data efficiency for AI workloads 

• Map out your data strategy upfront

Begin by clearly defining what data you need, where it will come from, how often it will be collected, and how it will be processed. Consider if data can be consolidated, stored using low-impact techniques, such as tape or other backup methods, or discarded if no longer necessary.  Offloading non-essential data to more energy-efficient storage methods can reduce power consumption. 

• Clean up before you start

Data efficiency in AI goes beyond just storing useful data. Data sets should be cleaned and optimised before training a model. Using raw, off-the-shelf data sets or repositories without minimising them results in unnecessary work and inefficiency. Cleaning data upfront ensures the model works more effectively and requires fewer resources.

• Get the training data set right

Data efficiency starts with an optimised data set for training, and using customer-specific data during model tuning helps further refine the model. By ensuring that data is as concise as possible from the start, you set a foundation for efficient processing throughout the entire AI lifecycle.

• Process data only once

Once data is processed for training/tuning, avoid redundant processing. Any additional training or fine-tuning should only occur on new data, minimising repeated energy-intensive operations.

Also Read: How a data-driven approach can optimise decarbonisation in the built environment

• Avoid data debt

Managing data is especially critical for AI workloads due to the massive volumes of data, including unstructured data. One of the key strategies for reducing the environmental burden of data is eliminating inaccurate, erroneous, out of date, or duplicated data. Like technical debt, data debt – where outdated or unnecessary data accumulates – can severely impact AI systems’ performance and sustainability.

• Location matters

Processing data as close to its source as possible minimises the energy required to move it. Optimising data movement reduces both the environmental and time-related costs, ensuring faster, greener AI operations.

As AI becomes integral to industries like manufacturing, logistics, and smart city initiatives, the need for more sustainable AI practices across Asia Pacific becomes more pressing. Singapore is already making strides in this area, with a focus on sustainable data centre innovation and initiatives like the . This approach is critical to ensuring AI systems can scale responsibly.

Asia Pacific’s growing dependence on AI-driven technologies presents a unique opportunity for the region to lead by example. Through initiatives that promote energy-efficient data management and more sustainable AI strategies, Singapore is positioning itself as a global leader in creating sustainable AI ecosystems.

To build a sustainable AI ecosystem in Asia Pacific, organisations must start with clean, lean data. As AI technologies become more embedded across industries, ensuring the data feeding into these models is optimised will not only help reduce energy consumption, but also foster innovation in a way that is more environmentally responsible.

For businesses in Singapore and the wider APAC region, prioritising data efficiency today will help ensure AI’s potential is fully realised without compromising the planet’s future.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. Share your opinion by submitting an article, video, podcast, or infographic.

Join us on Instagram, Facebook, X, LinkedIn, and our WA community to stay connected.

Image courtesy: Canva Pro

The post The hidden barrier to AI sustainability: Why clean data matters appeared first on e27.

Tashi Network, a company operating as the coordination layer for intelligent systems, has closed an oversubscribed funding round, aiming to empower trustless coordination for robots and AI agents.

The round was co-led by Blockchain Founders Fund (BFF) and Exponential Science Capital (ESC). The roster of participants also included Taisu Ventures and MN Capital.

The deal also attracted high-profile industry angels, including Gabby Dizon, co-founder of Yield Guild Games, and Wei Zhou, CEO of Coins.ph and former CFO of Binance.

Also Read: Blockchain and AI copyright: A revolution in digital rights management

The capital raised will be used to support rapid network growth and the company’s upcoming token launch on the Solana ecosystem.

Based in Singapore and California, Tashi focuses on replacing centralised coordination systems with a verifiable, distributed framework that allows machines to synchronise, validate, and settle actions in real time, without relying on central servers.

The company’s technology essentially transforms coordination itself into a form of currency; something that is measurable, rewardable, and tradable across the intelligent economy.

The startup’s experienced team of serial entrepreneurs has had multiple previous successful exits, including one sale valued at US$34 million to a company listed on the New York Stock Exchange. Amar Bedi, CEO of Tashi, previously held roles at tech giants Grab, Uber, and KPMG.

“The next computing revolution will allow trustless coordination among edge devices,” stated Bedi. He emphasised that Tashi’s core consensus technology, known as Vertex, enables the offering of “trust-preserving, global-scale coordination without any servers for the first time.”

Solving the centralisation problem

Tashi was explicitly designed to address existing industry challenges, as evidenced by the recent major outages experienced by companies like AWS. Current systems, including centralised clouds and global chains, often fail to deliver the instant, local coordination required by AI, robotics, and autonomous systems.

The company addresses this foundational problem by bridging decentralised physical infrastructure network (DePIN) with a novel approach to peer-to-peer consensus.

To understand how Tashi operates, imagine a massive orchestra composed entirely of robot musicians and AI conductors. Normally, they rely on one powerful, centralised maestro (a server) to tell them when and how to play. If that maestro has a cough or disappears, the entire concert stops.

Also Read: How AI and blockchain collaborate for a transparent Web3 future

Tashi, however, gives every single musician a small, independent mechanism to verify and synchronise their actions with everyone else instantly and securely. This means they can perform a complex symphony perfectly without any single conductor being in charge, ensuring the music never stops.

The firm is already demonstrating significant traction; it has built products targeting specific industries and currently boasts over 100 ecosystem partners and has secured early paying clients. Its DePIN already supports more than 50,000 nodes, which are run by the community.

The post Ex-Grab exec’s Tashi Network bags funding to kill AI’s centralisation problem appeared first on e27.

If you ask most people what mRNA is, they will likely scratch their heads and, if you’re lucky, draw a vague connection to Covid-19 shots. The acronym for ‘messenger ribonucleic acid’ entered the public consciousness during the pandemic as the foundation of preventive treatment against the virus.

How does it work? In simple terms, mRNA is a set of instructions your cells use to make specific proteins that help your body function and stay healthy. The Covid-19 vaccine is a well-known example of this technology in use, which essentially trains the body to combat the disease by stimulating an immune response.

Helping populations around the world protect themselves against one of the most virulent diseases to emerge in living memory is one way that physicians have used mRNA. Innorna, a biotechnology company specializing in mRNA engineering and lipid nanoparticle (LNP) delivery platforms, is advancing a potentially world’s first bivalent mRNA vaccine candidate against respiratory syncytial virus (RSV)—IN006, which targets both RSV-A and RSV-B—into clinical trials.

Innorna sees mRNA as adding flexibility to a vaccine in case of viral mutation. The bivalent nature of the vaccine means it can cover both RSV subtypes that currently exist. This design aims to provide broad-spectrum and durable protection.

A milestone in RSV prevention: IN006 completes phase 2 clinical study enrollment

In a significant step forward, Innorna has completed enrolment and vaccination in its Phase 2 clinical trial for IN006. This milestone marks a key step in the development of this innovative vaccine candidate, which is also recognized as China’s first domestically developed RSV vaccine to enter clinical trials.

The Phase 2 study is a randomized, double-blind, placebo-controlled trial conducted in China among healthy adults aged 60 and above. It is a critical step for dose optimization, broader population validation, and evaluating a booster shot for annual revaccination. This progress sets a solid foundation for subsequent Phase 3 efficacy studies.

“This Phase 2 clinical trial marks a critical step in validating IN006’s scientific hypothesis—delivering broad-spectrum, durable protection against RSV,” said Dr. Linxian Li, Founder and CEO of Innorna. “We remain committed to advancing the clinical development of this vaccine candidate efficiently. Our goal is to deliver safer, more effective RNA medicines to meet global public health needs.”

Also read: Cracking the code-switch: How a Hong Kong AI firm helps turn linguistic chaos into commercial clarity

Addressing a critical public health gap

RSV is a highly contagious virus that poses elevated risks to older adults, young children, immunocompromised individuals, and those with chronic conditions—potentially leading to pneumonia, respiratory failure, or death.

With no approved antiviral treatment and no authorized RSV vaccine in China, the need for effective prevention remains critical. IN006 represents a major step in public health innovation to fill this void.

Built on a proprietary technology platform

IN006 is built on Innorna’s proprietary pre-fusion F protein design, mRNA, and LNP platforms. Preclinical studies showed a favorable safety profile and strong humoral and cellular immune responses. In the preclinical cotton rat challenge study, IN006 provided effective protection against both RSV-A and RSV-B.

Innorna’s expertise in LNP enables it to serve as a platform licensor for vaccine and drug delivery. The company’s proprietary, rationally designed lipid library comprising over 6,000 chemically diverse ionizable lipids enables breakthroughs in mRNA vaccines and therapeutics.

Also read: Drawing the line on manual drudgery: Automation leader FJ Dynamics transforms unseen work with robotic precision

Leaping into unchartered territories to make medical miracles

Since forming Innorna six years ago, Dr. Li says the infrastructural support that Hong Kong Science and Technology Parks Corporation (HKSTP) provides has also been vital in taking the company to where it is today.

“We have received a lot of support from HKSTP, which has included both funding and other essential resources,” he explains. “They also help us connect with potential collaborators, which is very important for us because in the end, whatever groundbreaking treatments we may develop, we need to be able to commercialise them.”

In a field where preventive options are still limited, Innorna is applying big science to a pervasive viral challenge. The company’s work on IN006 aims to bring a powerful new tool to the global fight against RSV. This candidate has the potential to benefit a broad population and reshape the landscape of respiratory disease prevention.

—

Want updates like this delivered directly? Join our WhatsApp channel and stay in the loop.

This article is sponsored by HKSTP

We can share your story at e27 too! Engage the Southeast Asian tech ecosystem by bringing your story to the world. You can reach out to us here to get started.

Featured Image Credit: HKSTP, Innorna, Canva Images

The post Rewriting the future of RSV prevention: Innorna’s bivalent mRNA vaccine accelerates toward patients appeared first on e27.

In 2025, Indonesia’s startup ecosystem reached a defining moment. Sustainability took centre stage, investors renewed interest in green innovation, and communities grew more conscious of energy equity. Yet, for the clean energy sector, the heartbeat of Indonesia’s low-carbon transition, the pace of progress still lagged behind its potential.

The critical question is no longer if Indonesia can lead in renewable energy, but how quickly the ecosystem can bridge the gap between innovation and implementation.

Bridging the gap between innovation and implementation

Indonesia’s vast renewable potential spanning solar, hydro, and bioenergy remains one of the most promising in Asia. Yet, 2025 exposed persistent structural and systemic frictions: complex licensing procedures, uneven policy alignment between central and local governments, and a financing landscape that often undervalues early-stage climate ventures.

At Green Sphere Power Company, we experienced these bottlenecks directly. Our flagship initiative, a €2.5 million (US$2.7 million) renewable energy project with a 500kWh capacity, was designed to supply affordable and clean electricity to 500 households and 60 small businesses, schools, and healthcare centres in rural communities. The model demonstrated both scalability and impact. However, accessing consistent financing, navigating prolonged regulatory approvals and securing incentives for distributed generation posed real barriers to timely execution.

What held Indonesia’s clean energy startups back in 2025 was not a lack of ideas or ambition, but the ecosystem gap between innovation and capital readiness. Many renewable energy ventures were caught in a “pilot trap”, able to design technically viable solutions but unable to demonstrate financial bankability without early catalytic investment.

Venture investors still perceived clean-tech startups as high-risk due to the long payback periods and infrastructure-heavy models. As a result, founders had to rely on fragmented funding sources like grants, competitions, or private loans that were rarely synchronised with long-term sustainability goals.

Another major barrier was the shortage of technical talent. Indonesia’s renewable energy workforce remains underdeveloped, particularly in solar engineering, micro-grid design, and energy management. With regional competition from Vietnam, Malaysia and Singapore, local innovators often faced brain drain at a critical phase of growth.

Also Read: What new digital solutions mean for Indonesia’s F&B sector

Turning barriers into opportunities for 2026

If these gaps persist, Indonesia risks losing its competitive edge as Southeast Asia’s emerging clean-energy hub. But 2026 offers a unique opportunity for recalibration. The government’s renewed focus on green investment incentives, simplified renewable licensing, and integrated public-private partnerships could reshape the entire landscape.

To accelerate Indonesia’s transition from potential to progress, three strategic actions stand out:

• Mobilise blended finance: Combine public grants with private investment to derisk early-stage renewable projects like ours. A dedicated Green Innovation Fund could unlock millions in stalled clean-energy initiatives.
• Simplify permitting processes: Streamline national and regional regulatory frameworks to accelerate project approval timelines from months to weeks.
• Build technical capacity: Partner with universities and vocational institutes to train young engineers, entrepreneurs and technicians in renewable energy technologies.

We are actively contributing to this transformation by training climate entrepreneurs, helping them develop investment-ready project plans, and connecting them with investors who value sustainability alongside profitability.

Indonesia’s resilience in 2025 has laid the groundwork for renewal in 2026. The barriers of regulatory friction, fragmented finance, and talent scarcity can become catalysts for transformation if addressed collaboratively.

The next phase of Indonesia’s clean energy journey will not be defined by isolated innovation, but by ecosystem alignment—where policymakers, investors, and entrepreneurs move in sync toward a shared sustainability vision.

Indonesia doesn’t just have the potential to power its future; it has the opportunity to lead the region’s energy transition. What held us back in 2025 can be the very reason we accelerate in 2026.

The future of clean energy is not waiting for us to catch up. It’s waiting for us to lead.

—

Are you ready to join a vibrant community of entrepreneurs and industry experts? Do you have insights, experiences, and knowledge to share?

Join the e27 Contributor Programme and become a valuable voice in our ecosystem.

Image credit: Canva

The post Building Indonesia’s green momentum: What comes after 2025’s lessons appeared first on e27.

Manufacturing is changing fast. And one of the biggest shifts we’re seeing? Factories are starting to “see” for themselves.

That’s what CCTV-powered Vision AI does. It gives machines the power to understand what’s happening on the shop floor just by watching through CCTV cameras. Imagine security cameras that don’t just record footage but also think and react — spotting faulty products, noticing safety risks, or catching machine issues before they cause downtime.

In Southeast Asia, this trend is picking up fast. In just the first half of 2024, the region pulled in over US$30 billion in AI infrastructure investments.

So what does all this mean for manufacturing leaders on the ground? Let’s break it down.

Key trends

• Real-time quality checks: Traditionally, checking product quality meant human inspectors going through batches one by one. It’s slow, and errors slip through. But CCTV-based Vision AI changes that. It watches the production line 24/7 and instantly spots tiny defects like colour mismatches, cracks, or missing parts, before they move forward.
• Predictive maintenance: Machines break down when you least expect them to. But Vision AI can prevent that. By analysing live CCTV feeds, it can notice unusual movements, vibrations, or leaks in machines, early signs something’s about to go wrong.
• Safer workplaces: Safety lapses are expensive and dangerous. Vision AI can track worker behaviour on CCTV like checking if people wear helmets, gloves, and safety jackets. It can also alert managers instantly if someone enters a restricted zone or stands too close to heavy machinery.
• Data-driven insights: CCTV-based Vision AI systems don’t just watch — they collect data. This data shows where slowdowns happen, which processes create the most waste, and where productivity dips.
  

Challenges and barriers

Of course, adopting CCTV-based Vision AI isn’t all smooth sailing.

• High upfront costs: The tech isn’t cheap. Installing high-quality CCTV networks, training Vision AI models, and integrating them with existing systems costs a lot upfront. For small or mid-sized manufacturers, that can be intimidating.
• Data privacy concerns: CCTV cameras capture a ton of visual data, often including workers. So companies must follow strict data protection rules to make sure the footage is stored safely and used only for its intended purpose. Mishandling it could create legal risks.
• Need for skilled people: Vision AI systems need people who can maintain them, train models, and handle data. Many factories don’t have this talent in-house yet, and hiring or training new staff takes time.
• Change resistance: Not everyone will be thrilled about “AI watching them work.” Workers may worry about surveillance or job loss. It’s important for leaders to clearly explain that the tech supports them — not replaces them.

While CCTV-based Vision AI offers big benefits, it also needs careful planning, training, and clear policies to be successful.

Also Read: Enhancing cyber supply chain resilience: A vision for Singapore

Opportunities and the road ahead

Despite these hurdles, the opportunities are massive, especially for fast-growing manufacturing hubs in Southeast Asia.

• Fast scaling: Factories can grow operations without needing to hire and train a huge workforce. Vision AI can handle quality checks, track safety, and analyse productivity, letting teams focus on creative and complex tasks.
• Cost savings: Less downtime, fewer product defects, and fewer workplace accidents directly save money. Companies that adopt Vision AI early can become more competitive by lowering waste and speeding up output.
• Sustainability wins: CCTV-powered Vision AI helps spot energy waste, reduce material scrap, and improve resource usage. That makes operations more eco-friendly.
• Staying ahead of the curve: With the global market shifting toward Industry 4.0, companies using Vision AI now will be ahead of the curve. Early adopters will build smarter, safer, and more flexible factories — ready to handle future challenges.
  

The future is clear: CCTV-based Vision AI isn’t just an add-on. It’s becoming the nervous system of modern factories — watching, learning, and guiding production in real time.

Conclusion

Manufacturing is entering a new era where CCTV cameras don’t just watch, they think.

CCTV-powered Vision AI is helping factories catch defects instantly, prevent machine failures, keep workers safe, and improve efficiency — all at once.

Yes, it comes with challenges like cost, privacy, and training needs. But the long-term benefits far outweigh the risks.

For manufacturers in fast-growing regions like Southeast Asia, now is the time to explore this shift. Because in the coming years, smart eyes on the factory floor won’t be a luxury — they’ll be a necessity.

—

Are you ready to join a vibrant community of entrepreneurs and industry experts? Do you have insights, experiences, and knowledge to share?

Join the e27 Contributor Programme and become a valuable voice in our ecosystem.

Image credit: Canva

The post How CCTV-based vision AI is transforming manufacturing appeared first on e27.

Sea Limited, the Singapore-based consumer internet giant, has released its Q3 2025 results, solidifying its return to high-growth, bottom-line profitability and reporting a stunning surge in net income.

However, a closer inspection of the financial details reveals that the rapid expansion of its digital financial services segment, Monee, is accompanied by a sharp acceleration in credit risk provisioning.

The overall financial momentum is undeniable, with the company reporting total GAAP revenue of US$6 billion, marking an increase of 38.3 per cent year-on-year (YOY) from US$4.3 billion in Q3 2024. Total net income rocketed to US$375 million, soaring 144.6 per cent YOY compared to the US$153.3 million recorded in the corresponding period last year.

Also Read: Sea posts 418% profit jump as Shopee, Monee, Garena fire on all cylinders

Total adjusted EBITDA stood at US$874.3 million, up 67.7 per cent YOY.

Digital entertainment and e-commerce drive profit surge

The company’s three core businesses — Garena (digital entertainment), Shopee (e-commerce), and Monee (digital financial services) — all contributed robustly to the group’s performance.

Digital Entertainment (Garena): This segment delivered exceptional results, with CEO Forrest Li stating, “Garena has delivered another stellar quarter. Bookings were up 51 per cent year-on-year, making it our best quarter since 2021.”

• Bookings reached US$840.7 million, increasing by 51.1 per cent YOY.
• Paying users grew 31.2 per cent YOY to 65.9 million, resulting in a paying user ratio of 9.8 per cent (up from 8.0 per cent in Q3 2024).
• Adjusted EBITDA for the segment was US$465.9 million, up 48.2 per cent YOY. This success was largely anchored by “two high-impact campaigns: Squid Game and NARUTO SHIPPUDEN Chapter 2” for Free Fire.

E-commerce (Shopee): Shopee cemented its profitability turnaround, posting an adjusted EBITDA of US$186.1 million, a staggering increase of 440.1 per cent from US$34.4 million in Q3 2024.

• GAAP revenue for the segment hit US$4.3 billion, up 34.9 per cent YOY.
• Core marketplace revenue, which consists of transaction-based fees and advertising revenues, grew by 52.8 per cent YOY to US$3.1 billion.
• Nuance in e-commerce: While core fees surged, value-added services revenue (primarily logistics-related) saw a decline of 5.7 per cent YOY to US$723.6 million. The company attributed this decrease to “higher revenue net-off against shipping subsidies”.

The unavoidable risk of rapid credit growth

While segment growth narratives were overwhelmingly positive, the most dramatic increase in expenditure was found in the provision for potential bad debts, highlighting the structural risk associated with the booming credit business.

Digital financial services (Monee): This segment remains the fastest growing by revenue percentage.

• GAAP revenue reached US$989.9 million, marking a robust 60.8 per cent YOY growth, primarily driven by the growth of the credit business.
• Consumer and SME loans principal outstanding grew significantly, up 69.8 per cent YOY to US$7.9 billion as of September 30, 2025.

The underlying nuance: Credit provision surge

Despite the growth, the provision for credit losses saw a massive increase of 76.3 per cent, jumping from US$212 million in Q3 2024 to US$373.8 million in Q3 2025. This provisioning expense grew significantly faster than the segment’s adjusted EBITDA, which was up 37.5 per cent YOY to US$258.3 million.

Also Read: Sea Limited’s 2024 results: A deep dive beyond the headlines

Sea Limited noted that the non-performing loans (NPLs) past due by more than 90 days remained stable at 1.1 per cent of the total loan principal outstanding (including on-book and off-book loans). While the NPL ratio suggests stability, the sheer scale of the 76.3 per cent increase in provision expense signals that the substantial expansion of lending activities, particularly the US$7.9 billion in principal outstanding, inherently carries rapidly increasing absolute risk exposure. This is a critical detail in gauging the long-term sustainability and quality of the digital finance segment’s profits.

Playing down investment in the future

Another detail that provides insight into Sea’s current strategy is the allocation of operating expenses.

Total operating expenses grew by 28 per cent overall. However, expenses related to future innovation were curtailed:

• Research and development expenses actually decreased by 5.2 per cent, falling to US$286.3 million in Q3 2025.
• In contrast, sales and marketing expenses surged by 30.9 per cent to US$1.2 billion, demonstrating a clear prioritisation of immediate market capture and revenue acceleration over investment in future technological development during this period. This shift is particularly evident in the Digital Financial Services segment, where sales and marketing expenses soared by 140.7 per cent.

In summary, while Sea Limited’s Q3 results rightly celebrates a decisive return to high profitability, underscored by record Garena performance and a Shopee turnaround, the sharp 76.3 per cent jump in credit loss provisions alongside a reduction in R&D spending suggests the company is aggressively pursuing current period growth and profitability in Southeast Asia, even if it means ramping up balance sheet risk and marginally slowing future technology investment.

The post Sea Limited roars back to profit, yet credit loss provisions flash warning signs appeared first on e27.

The landscape of venture capital exits is undergoing a massive reassessment, particularly concerning merger and acquisition (M&A) activity.

While strategic acquisitions remain a critical exit route, the market has shifted dramatically towards smaller deals characterised by higher buyer caution and increased structural complexity. This trend holds significant implications for startups across Southeast Asia (SEA) aiming for major acquisition events.

The insights from the comprehensive study State of Exits 2025: From Alpha to Omega by RETVRN Research, reveal a stark reality in the M&A world: scale has become elusive for most. Despite modest improvements in IPO activity, particularly with 13 US-based venture-backed companies going public at US$1 billion-plus valuations in 2025 YTD (as of July 2025), compared to just eight for the entire year of 2024, the overall M&A picture remains highly constrained.

Also Read: The new exit reality: How secondary deals became the lifeblood of venture capital

The most telling data point reinforcing the “reality gap” is the distribution of transaction values. The study confirms that 96 per cent of M&A transactions are valued below US$500 million. Furthermore, a significant portion of these deals is concentrated at the lower end of the spectrum, with 70 per cent of all M&A transactions valued under US$100 million.

This statistic paints a clear picture: mega-exits remain rare, and the majority of liquidity events fall into the realm of small to mid-sized strategic acquisitions.

Buyer selectivity and value growth amid volume decline

Global M&A volumes declined by 9 per cent in the first half of 2025. However, counterintuitively, deal values increased by 15 per cent during the same period.

This contradiction is highly revealing; it signifies a market that has become exceptionally selective, consistently favouring only the highest-quality assets. Acquirers are deploying large sums for proven, critical technologies, but are pulling back on speculative or merely average opportunities.

For founders, particularly in SEA, where large regional conglomerates or international players are the typical buyers, this means the bar for being considered a “high-quality asset” has never been higher. The market is no longer forgiving of volatile performance or unproven unit economics.

The pervasiveness of structured deals

Perhaps the most structural change affecting M&A negotiation is the rise of structured deals. The shift reflects profound caution on the part of buyers and a corresponding willingness by sellers to share risk. This is fundamentally altering how M&A transactions are negotiated and valued.

A staggering 73 per cent of deals now include extensive earnout provisions. Moreover, an average of 42 per cent of the total consideration is contingent on future performance metrics. This mechanism ensures that the buyer pays a significant portion of the price only if the acquired company meets predefined milestones after the transaction closes.

The commitment period for sellers has also extended considerably. The average earnout period has stretched to 3.2 years, up significantly from 2.1 years in 2020. This means founders and key staff are now tied to the acquiring entity and its performance metrics for a much longer duration to realise the full transaction value.

Also Read: Secondaries take centre stage: How VCs are navigating the exit drought

Founders must prepare for this reality by ensuring their internal operational excellence is impeccable, with reliable forecasts that deliver consistent results (within a margin of plus or minus 10-15 per cent accuracy).

Valuation compression and the SaaS reality check

The market correction following the peak years has had a profound impact on sector valuations, particularly in enterprise SaaS. RETVRN Research notes that valuation multiples for Enterprise SaaS have experienced severe compression, declining from peak levels of 15 times revenue to a median multiple that has now stabilised at 7.0 times current run-rate annualised revenue.

While this stabilised multiple is consistent with historical norms, it represents a dramatic correction from the 2021 peaks. Peak multiples were reached in Q4 2021 at 12.8 times revenue for SaaS companies, before hitting a trough in Q3 2023 at 3.2 times revenue—a 75 per cent decline. The subsequent recovery pattern shows gradual stabilisation in the 6-8 times revenue range.

For bootstrapped companies, the median average is even lower, at 4.8 times revenue, while equity-backed companies average 5.3 times revenue. These figures confirm that while the market is recovering, the era of exuberant, growth-at-all-costs valuations is over. The median valuation multiple has stabilised, demanding disciplined financial performance from all founders.

The early exit strategy

The data indicates that planning for an exit must begin much earlier than many founders currently assume. Over 60 per cent of acquisitions happen at or before the Series A stage. Specifically, 47 per cent of acquisitions analysed in 2025 were Seed-stage acquisitions.

Also Read: What did we learn from failing to raise VC funding?

The majority of founders are always closer to an exit than they realise, but a lack of exit planning often exposes them to mediocre exit outcomes. Given the dominance of small-to-mid-sized deals and the prevalence of earnouts, achieving a premium valuation now relies entirely on early strategic alignment, clean operational data rooms, and proactive cultivation of strategic relationships 18–24 months before the intended exit. This focused preparation is the only way to successfully navigate the highly selective, structure-heavy M&A landscape defined by caution and a firm grip on reality.

The post Founders face a brutal new reality: Tiny exits, tougher buyers, endless earnouts appeared first on e27.