Singapore’s AI national strategy has received its most significant tune-up since the launch of NAIS 2.0 in December 2023. Describing the update as a “double-click rather than a system reboot,” Minister for Digital Development and Information Josephine Teo announced 10 refreshed priorities at the ATxSummit on May 20, underscoring the government’s intent to accelerate rather than reinvent its AI agenda.

The refreshed priorities follow the establishment of the National AI Council in February 2026, chaired by Prime Minister Lawrence Wong, which was set up to provide strategic direction for Singapore’s AI agenda. The update is organised around three focus areas: deepening sectoral and public sector transformation, mainstreaming AI adoption across the broader economy, and cementing Singapore’s position as a regional AI hub.

A central pillar of the plan is the launch of National AI Missions in advanced manufacturing, financial services, connectivity, and healthcare — four industries that are critical to the country’s economic backbone. Together, those four sectors contributed roughly 40 per cent of Singapore’s GDP in 2025. The strategy envisages AI being embedded more deeply across government agencies as well, with the aim of accelerating public sector transformation and improving citizen services.

Workforce development features prominently. The National AI Impact Programme targets 10,000 SMEs for meaningful AI adoption, while the Champions of AI programme offers more targeted support for enterprises seeking to go further. Broad-based capability-building and what the strategy terms “AI bilingual talent” — professionals fluent in both domain expertise and AI application — are positioned as foundational requirements for the transformation ahead.

Also Read: Singapore lands OpenAI’s first lab outside the US with US$225M commitment

On the infrastructure front, Singapore will expand local research compute capacity from 2026 through the National Supercomputing Centre’s ASPIRE 2B supercomputer, as part of a planned national advanced compute, AI, and scientific computing platform. A Digital Infrastructure Act is also expected to be tabled in Parliament to set baseline sustainability standards for data centres.

Internationally, Singapore is leaning into its role as a convener. Minister Teo was candid about the country’s constraints, noting that Singapore’s domestic market alone may not warrant the level of attention it receives, but that its value lies in the global networks it is connected to and its track record in trusted technology adoption.

Business analysis: Opportunity tempered by execution risk

From a business perspective, the updated strategy presents a compelling and largely coherent value proposition. The targeting of four high-GDP sectors with dedicated AI Missions gives multinationals and local enterprises alike a clearer signal of where government support, regulatory clarity, and talent pipelines will converge.

The commitment of more than S$1 billion to public AI research and talent development from 2025 to 2030 — announced earlier this year — provides meaningful financial scaffolding for the private sector to build upon. NVIDIA’s new Research Lab in Singapore and the Punggol Digital District are early markers that global tech players are already responding to Singapore’s hub ambitions.

Also Read: Why you should be hiring humans when others are hiring AI agents

Yet the strategy is not without its risks. The sheer breadth of the 10 refreshed priorities — spanning compute, data governance, workforce capabilities, international partnerships, and sectoral transformation — raises legitimate questions about execution capacity and coordination across agencies.

Smaller enterprises, in particular, may find the SME-focused programmes insufficient to keep pace with the pace of AI disruption, especially as global competitors ramp up investment at scale. Singapore’s acknowledged constraint of a small domestic market also means that the hub ambitions are ultimately contingent on sustained foreign investment and geopolitical stability — factors that lie well beyond any national strategy document. For businesses watching closely, the direction is right; the proof will lie in delivery.

The post Singapore’s AI National Strategy gets a sharp refresh with business ambitions front and centre appeared first on e27.

Quantum themes are now everywhere—from multiverse plots to Schrödinger’s cat, the famous paradox of a cat that is both alive and dead until observed. This surge in storytelling isn’t just cultural curiosity. Experts see it as a sign that imagination is beginning to meet scientific feasibility.

When Sci-Fi becomes signal

The trajectory echoes that of artificial intelligence. AI appeared in fiction for years, but only reshaped industries after deep learning breakthroughs in 2012. Quantum computing may be approaching a similar tipping point.

Until recently, quantum ideas lived mostly in theory. But researchers are now building real systems—stabilising qubits and testing early quantum algorithms.

As science catches up to storylines, analysts note that public imagination often signals where real progress is emerging. For those watching closely, quantum’s rise in fiction could be more than a coincidence—it could be the earliest sign of real-world transformation.

A new kind of computation

Quantum computers differ fundamentally from classical machines. While traditional systems handle tasks step by step, quantum computers explore many possibilities at once—thanks to quantum phenomena like superposition and entanglement.

One expert likens it to a scene from Avengers, where Doctor Strange scans millions of futures simultaneously to find the best path forward. That’s essentially how quantum systems approach complex problems—by evaluating countless outcomes in parallel.

This makes them especially suited for challenges that overwhelm classical systems, such as:

• Cracking next-generation encryption
• Optimising vast logistics networks
• Simulating molecular interactions in drug or material discovery

Quantum isn’t here to replace everyday computing. But for specific high-complexity problems, it represents a fundamentally new—and more powerful—computational model.

Where quantum will hit first

Quantum computing is expected to make its earliest impact in industries where computational complexity is high and financial upside is significant.

Key early applications include:

• Pharmaceuticals: simulating molecular interactions to accelerate drug discovery
• Advanced materials: designing new compounds or batteries at the atomic level
• Finance: optimising asset portfolios, particularly in ETFs and derivatives

Take ETF construction, for example. Selecting the ideal combination of dozens of assets involves combinatorial optimisation—a task that becomes exponentially harder as the number of variables increases. While AI tools help, classical systems struggle beyond a point. Quantum computers, by evaluating multiple combinations simultaneously, offer a clear advantage.

Also Read: Navigating Asia’s business boom: The quantum leadership advantage

In the short term, industries that combine high complexity with high value potential are best positioned to adopt quantum solutions—because the benefits justify the infrastructure investment.

Early wins in the quantum race

While today’s quantum computers remain in the early stages—most with only a few dozen usable qubits—they are already beginning to show practical value in select domains. Many available qubits are still dedicated to error correction, reflecting the sensitivity of current hardware.

Yet despite these limits, meaningful use cases are emerging.

Notable early applications include:

• Drug discovery: simulating molecular behaviour at quantum levels
• Advanced materials: modelling atomic interactions for next-gen compounds
• Finance: improving asset rebalancing strategies in complex portfolios
• Logistics: optimising large-scale routing problems that scale exponentially

These are all areas where classical systems struggle as complexity increases.
Even with today’s constraints, quantum systems are starting to outperform traditional methods in narrow but high-impact scenarios.

The technology may still be maturing, but its real-world value is no longer theoretical—it’s beginning to take shape.

Why quantum-AI hybrids matter

Quantum–AI hybrid computing is drawing growing attention as a practical way to extract early value from quantum systems. Rather than replacing classical computing, the hybrid model assigns different parts of a task to the most suitable processor.

• Classical computers or AI handle large-scale, repetitive calculations
• Quantum systems tackle tasks involving simulation, optimisation, or quantum-specific modelling
• Cloud platforms or machine learning layers integrate and interpret the combined outputs

This division of labour leverages the strengths of each architecture, delivering faster, more efficient results than either could alone. Experts see hybrid models as the most viable short-term strategy—not only technically, but also commercially and operationally—to scale quantum’s impact without waiting for perfect hardware. 

Rethinking the internet for a quantum era

If quantum computers reach commercial viability, today’s internet security architecture will require a complete overhaul. Most current encryption methods—used in banking, e-commerce, communications, and authentication—are vulnerable to quantum algorithms capable of breaking them. This wouldn’t just call for software updates; it would demand a structural redesign of global digital infrastructure.

Experts describe it as a foundational shift, not a technical patch. That said, the transition is expected to unfold gradually over the next decade, giving rise to quantum-resistant cryptographic standards and long-term planning by governments and enterprises.

For infrastructure providers and investors, the key is timing: anticipating when and how to adapt before disruption becomes inevitable.

Korea’s quantum edge: Beyond hardware

While Korea has produced notable quantum researchers, including one of IonQ’s co-founders, full-scale hardware development remains concentrated in global hubs like the US, where companies such as IBM, Google, and IonQ lead in capital and infrastructure.

Instead, Korea is gaining ground in quantum-resilient infrastructure, particularly in quantum-safe cybersecurity. A standout example is SK Telecom, which acquired Swiss-based ID Quantique—a global leader in quantum key distribution (QKD)—and later entered a strategic partnership with IonQ.

Also Read: Horizon Quantum CEO on the Singapore advantage in starting a quantum computing company

This positions Korea to lead in quantum-proof security systems, a field likely to reach commercial scale well before universal quantum computing becomes mainstream.

Experts draw parallels to the early AI wave (circa 2014–2015), when Korea didn’t build foundational frameworks but found success through application-level innovation. Similarly, Korea’s future in quantum may lie in industry-specific algorithms, secure infrastructure, and applied software—not hardware.

How big tech is positioning for quantum leadership

Major tech companies are taking two main paths toward quantum computing: in-house development and strategic partnerships or acquisitions.

• IBM and Google have pursued full-stack integration from the outset—developing quantum hardware, software tools, and embedding quantum capabilities directly into their cloud platforms. They remain the most vertically integrated players in the field.
• Microsoft and Amazon initially focused on enabling quantum access through their cloud ecosystems, partnering with startups to provide tools via Azure and AWS. But as the commercial potential grows, both are moving toward greater internal control.

Recent shifts include:

• Microsoft’s launch of proprietary tools through initiatives like MyOrionow
• Reports of Amazon collaborating with or acquiring startups such as OxenT to build its own quantum stack

This signals a broader trend: big tech is transitioning from collaboration to ownership, aiming to secure key positions as quantum computing moves from theory to viable markets.

Quantum investing: Echoes of the deep learning era

Some early-stage investors see clear parallels between today’s quantum computing landscape and the deep learning boom of the mid-2010s.

In 2012, deep learning began to show real promise. By 2013–2014, major tech firms were investing heavily. During that wave, investors backed AI startups that later went public after a 7-year growth cycle.

Quantum computing now appears to be following a similar arc:

• Foundational research is maturing
• Big tech is entering aggressively
• Use cases are emerging in finance, pharma, and cybersecurity

Many investors haven’t placed direct bets yet, but they’re watching closely.
Angel investors may need a 10-year horizon, while VCs could see returns within three to five years years.

The consensus among early movers? This is the beginning of a new wave—and the time to position is now.

The quantum era won’t arrive overnight—but once it’s here, it will move fast. Those who act early won’t just adapt to the future—they’ll help shape it.

Special thanks to Dr. Jihoon Jeong, General Partner of Asia2G Capital, for his valuable contributions to this article.

—

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 Quantum’s inflection point: Why the smart money is watching now appeared first on e27.

AI adoption across Southeast Asia is accelerating, and with the edge AI market expected to reach US$66.47 billion by 2030 (21.7 per cent CAGR), organisations are moving quickly from pilots to embedding automation in core operations. Agentic AI is expanding what can be automated, prompting enterprises to reassess their technology foundations amid growing regional expansion and rising demands on infrastructure design, visibility and control.

For manufacturers, diversification often replaces a mature ecosystem with a fragmented, cross-border supply chain. Navigating differing regulations and dispersed suppliers deepens reliance on public cloud and hyperscaler AI services, turning infrastructure choices into strategic decisions that determine performance, reliability and resilience.

At the same time, enterprises are shifting from traditional data centres to hybrid control planes. Intelligent edge devices now span factories, clinics and shop floors, unlocking new automation gains but also multiplying attack surfaces. This underscores the need to secure AI workloads consistently across distributed environments and build scalable, resilient architectures.

The AI-cybersecurity arms race

The barrier to entry for sophisticated cyberattacks has collapsed. Agentic AI allows attackers to automate reconnaissance, identify vulnerabilities and scale targeted attacks without deep expertise. Threats that once required manual skill can now be executed by prompting an AI agent to map weaknesses and exploit them efficiently.

Meanwhile, supply chain diversification has fractured defensive perimeters. Thousands of devices and sensors across dispersed facilities now form a sprawling attack surface. As business speed increases, attacks move at the same pace. Manipulative social engineering, deepfake voice impersonation and automated phishing campaigns overwhelm human analysts and exploit the weakest links.

Also Read: How an AI cybersecurity company harnesses the power of AI for optimal business performance

This raises a critical question: how can organisations detect and neutralise AI-enabled threats quickly enough to prevent meaningful damage?

The quantum computing effect

Quantum computing introduces another layer of urgency. As organisations expand their digital borders into fragmented environments, attackers gain conditions to automate and accelerate intrusions. One threat has become especially concerning: “harvest now, decrypt later”. Attackers can steal encrypted data today, store it and wait until quantum systems can break the underlying cryptography. Health records, intellectual property and long-term customer data could become liabilities once decrypted.

This makes the migration window critical. Organisations have limited time to upgrade cryptographic systems before quantum technologies render them vulnerable. Upgrading at scale – discovering dependencies, securing keys and deploying post-quantum algorithms across many systems takes years. If migration takes five years and quantum capability arrives in ten, the clock is already ticking.

Visibility complicates matters. Many enterprises lack a full inventory of keys, certificates or hard-coded encryption calls. Manual audits are slow and incomplete. AI-powered code scanners can accelerate discovery, map quantum-susceptible components and guide modernisation. AI can also detect subtle data exfiltration patterns and deploy countermeasures such as injecting fake data to neutralise stolen datasets.

Compliance will tighten across Asia

Regulators are tightening supply chain mandates and raising expectations for cybersecurity maturity. Japan’s Ministry of Economy, Trade and Industry (METI), will introduce its Cybersecurity Measures Evaluation System for Strengthening Supply Chains in 2026. Similarly, South Korea is strengthening cybersecurity oversight and Hong Kong’s Protection for Critical Infrastructures (Computer Systems) Bill, effective 2026, imposes stronger obligations on organisations to modernise defences.

Compliance is no longer a checkbox exercise — it is a strategic imperative tied to operational resilience and competitive readiness.

Data-heavy industries, look out

Healthcare illustrates the stakes. When sensitive data flows across cloud systems, hospitals and connected devices, even a minor breach can trigger cascading disruption. Similar vulnerabilities appear in manufacturing, logistics, finance and retail, where interconnected digital ecosystems amplify the impact of AI-driven threats.

Also Read: Unchecked shadow AI poses a major cybersecurity risk for 2026: Exabeam

A realistic scenario: an attacker scrapes public data to profile a medical professional, generating a cloned voice and calling the IT help desk to reset authentication credentials. Once inside, the attacker can move laterally and quietly. AI-powered defences are essential because they detect behavioural anomalies — unfamiliar browser fingerprints, impossible travel events or unusual directory access — rather than relying on malware signatures.

How enterprises can stay ahead

• Correlate telemetries at scale: Organisations can improve detection accuracy by correlating telemetry across networks, devices and applications. This uncovers hidden anomalies designed to evade traditional tools. Proactive red-teaming of AI models uncovers vulnerabilities such as data poisoning or manipulation. Explainable AI techniques support forensic analysis by showing why alerts were generated.
• Enforce data provenance and sanitisation: Security begins at the data layer. Organisations should validate data at every ingestion point and prevent modified or corrupted inputs from entering critical systems. Immutable ledgers or blockchain mechanisms ensure trusted provenance and integrity for high-assurance pipelines.
• Address the human element and “shadow AI”: Cybersecurity awareness must extend to all staff. Shadow AI – unvetted tools in daily workflows – poses a growing risk. Core hygiene practices such as least privilege access, multi-factor authentication and granular role-based controls remain essential. Training helps staff recognise modern risks, including seemingly harmless third-party AI tools that could execute tasks autonomously on corporate networks.

The winners in this new era will be those who treat AI security as a strategic advantage, not an afterthought. Building resilience at machine speed requires more than technology—it demands a mindset shift towards dynamic, multi-layered defence. In Southeast Asia’s AI-driven economy, confidence will belong to enterprises that synchronise innovation with security, turning risk into a competitive edge.

—

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.

Enjoyed this read? Don’t miss out on the next insight. Join our WhatsApp channel for real-time drops.

Image credit: Canva

The post AI adoption in Southeast Asia: Balancing automation gains with the rising threat of cyberattacks appeared first on e27.

Secai Marche, a Tokyo-headquartered startup operating a farm-to-table fresh food distribution platform across Southeast Asia, has closed an additional financing round led by NTT Docomo Ventures and Synexia Ventures.

As per the agreement, the startup has entered into a strategic partnership with NTT Data to digitise invoicing and payments for Malaysia’s HORECA (hotel, restaurant, and catering) sector.

Also Read: Japanese logistics giants bet on Secai Marche’s cold-chain network vision

The deal marks a shift in Secai Marche’s product roadmap: from pure logistics and marketplace services to embedding payments, invoicing, and financial products, such as buy now, pay later (BNPL), supply chain finance, and microloans for farmers and small producers. The move targets long-standing inefficiencies in Malaysia’s food procurement workflows, where manual invoicing and payment processes create back-office burdens and working capital headaches across the supply chain.

Why payments matter for fresh food distribution

Secai Marche’s platform connects producers directly with restaurants, retailers and other buyers, handling ordering, logistics and market development. The company argues that while physical distribution has seen incremental improvement, the transactional layer (invoicing, payment reconciliation, and financing) remains stubbornly analogue in many Southeast Asian HORECA markets. That gap not only slows operators but also leaves smallholder farmers exposed to irregular cash flows and delayed payments.

“Procurement, invoicing and payment workflows remain highly analogue, leaving substantial inefficiencies that can constrain business growth,” Secai Marche’s representative director said in a statement. The company plans to integrate NTT DATA’s payment and invoicing solutions into its marketplace to provide a unified service that covers procurement through to settlement.

For restaurants, automated invoicing and online payments can reduce human error, speed reconciliation and improve visibility into cash positions. For producers, digitised accounts payable (AP) flows can shorten receivable cycles and open the door to underwriting based on transaction histories rather than traditional collateral, a key point if Secai Marche’s BNPL and supply chain finance ambitions materialise.

NTT DATA’s role: tech, payments and ecosystem

NTT Data will supply payment and invoicing technology and collaborate on integrating those capabilities into Secai’s marketplace. Shinichiro Nishikawa, head of NTT Data’s Global Payments & Services Division, framed the collaboration as more than operational efficiency: “We believe that enhancing financial services through the utilisation of payment and invoice data can help improve companies’ capital efficiency and create new access to finance.”

That line signals a common fintech playbook: digitise flows, capture transaction data, then overlay credit and liquidity products. For NTT Data, the partnership aligns with broader ambitions to leverage enterprise payment data for financial services across markets. For Secai Marche, it could mean moving from being a logistics layer to becoming a finance-enabled supply chain platform.

Venture backers see regional scale

NTT Docomo Ventures, one of the new investors, said it backed Secai because the startup addresses fundamental pain points for farmers and HORECA operators and because the NTT Group’s assets can accelerate scale. “We have high expectations for SECAI MARCHE’s growth into a platform that connects Southeast Asia and, ultimately, the world,” Yuma Kotake, Director at NTT Docomo Ventures, said.

Also Read: Secai Marche cultivates US$6M to build a fresher, smarter food ecosystem in SEA

Synexia Ventures is also making Secai Marche its inaugural portfolio investment. “Secai Marche has built a unique position in Southeast Asia’s farm-to-table fresh produce supply chain,” its MD Kuan Hsu said, signalling confidence that the combined JV with NTT DATA could unlock additional value.

Putting finance on top of perishables is not a trivial pursuit. Fresh produce supply chains operate on thin margins and tight timing constraints; payment products must be reliable, low-friction and closely integrated with logistics and invoicing data to succeed.

A practical case: how integration could change day-to-day operations

Today, many restaurant operators in Malaysia still receive paper invoices or spreadsheets, then manually approve payments and reconcile bank transfers. For popular eateries operating on slim margins, delayed supplier payments and opaque receivable cycles create unpredictability, forcing either precautionary cash buffers or frequent short-term borrowing.

Integrated invoicing and payments would let restaurants approve digital invoices, initiate payments from a single dashboard, and view real-time payment status. From the supplier side, producers could see when payments will arrive and, eventually, choose to monetise upcoming invoices through supply chain finance or BNPL arrangements underwritten by the platform using its transaction data.

By centralising procurement-to-payment flows, Secai Marche can also reduce reconciliation costs, a non-trivial overhead for SMEs that currently spend significant time on bookkeeping. This could be an attractive sell to restaurant chains and medium-sized caterers, which value predictable cash conversion cycles as much as timely deliveries.

Risks and execution challenges

The plan is ambitious and faces several hurdles. First, regulatory frameworks governing payments, lending, and data use vary across Southeast Asia; navigating Malaysian rules alone will require careful compliance and possibly local financial services licences. Second, underwriting credit to farmers and small vendors is inherently risky; default management and fraud prevention will be crucial. Third, customer adoption requires trust — operators and producers must be convinced that the platform’s financial services are reliable, affordable, and tailored to their cash flow patterns.

Moreover, embedded finance models often need scale to make economics work. Transaction volumes must be sufficient to justify credit exposure and to provide the data richness required for risk models. Secai Marche will need to expand its transaction footprint in Malaysia quickly while ensuring unit economics remain viable.

Strategic rationale and regional ambitions

For Secai Marche, integrating payments is a logical extension of its core marketplace. Data from orders, deliveries and invoices provides a relatively rich signal for credit assessment compared with alternative data sources. If executed well, the company could capture more of the value chain — from order placement to payment collection and financing — rather than just taking a cut on logistics or marketplace fees.

Investors’ emphasis on regional scaling suggests Secai Marche may replicate the model beyond Malaysia. However, the company will be tested first on the ground: convincing HORECA operators to switch from manual, familiar processes to an integrated, digital service and proving that financial products can actually reduce friction rather than add complexity.

Looking ahead: BNPL, supply chain finance and microloans

Secai Marche explicitly mentioned plans to explore BNPL for procurement, supply chain finance and microfinance for farmers, using accumulated transaction and invoice data to underwrite loans. Microfinance to upstream producers is a particularly appealing public-good narrative: it could stabilise upstream supply, improve quality and create longer-term relationships between producers and buyers.

Also Read: Secai Marche wins US$4M grant from Japan govt. to transform farm-direct e-commerce in SEA

But turning transaction history into credit access requires robust risk models and often access to capital or third-party underwriters. The partnership with NTT Data and backing from NTT Docomo Ventures may help here by opening institutional channels and technology resources, but execution will require both engineering and prudent financial risk management.

Bottom line

Secai Marche’s fresh capital and strategic partnership push the company into the crowded but potentially lucrative territory of embedded finance for supply chains. Its success will hinge on execution: building seamless integrations that materially reduce back-office friction, proving credit products that demonstrably help small producers and buyers, and navigating regulatory and operational risks in Malaysia before scaling across Southeast Asia.

If the farm-to-table startup can convert transaction data into reliable finance at scale, it could become a critical infrastructure layer for the region’s perishable food supply chains, but the path from marketplace to finance provider is littered with complexity. The new funding and ties to NTT Group assets give it a better shot than many, but much still depends on adoption, unit economics and risk control in the months ahead.

The post Secai Marche raises fresh capital to stitch payments into Malaysia’s HORECA supply chain appeared first on e27.

As a founder, one of the hardest lessons I’ve had to learn is this: You can outsource tasks, but you cannot outsource ownership.

Not to AI. Not to agencies. Not to communities. Not even to people genuinely trying to help you succeed.

And strangely enough, I didn’t learn this lesson from a failed product launch or a difficult investor meeting.

I learned it from people with potential.

Over the years of building businesses, communities, and founder ecosystems, I’ve met many individuals who were creative, intelligent, and full of ideas. Some of them were vocal, charismatic, and clearly capable of building something meaningful.

The potential was obvious.

But potential and ownership are not the same thing.

That distinction matters far more than most people realise.

Ideas are common, ownership is rare

One pattern I kept noticing was how easy it was for people to get excited about possibilities.

A new business idea. A personal brand. A community initiative. An AI tool. A collaboration opportunity.

In the early stages, energy is everywhere. Conversations are exciting. Ideas flow endlessly. Everyone feels inspired.

But the moment friction appears, things change.

Some people pause. Some people wait. Some people retreat. Some people start looking externally for reassurance, validation, or permission to continue.

Founders do not have that luxury for very long.

Because building anything meaningful requires the ability to continue moving even when things become uncomfortable, uncertain, or inconvenient.

Also Read: The shadow ledger: Why AI governance is the new architecture of brand trust and enterprise revenue

That is where ownership begins.

And over time, I realised something uncomfortable myself: Many people want the outcome of entrepreneurship without fully accepting the responsibility that comes with it.

They want growth without consistency. Visibility without vulnerability. Momentum without initiative.

Most importantly, they want transformation without ownership.

The founder trap: Caring more than the other person

I naturally enjoy helping people build.

It’s one of the reasons I created founder communities, educational programmes, and AI-powered systems in the first place. I genuinely love seeing people gain confidence, clarity, and momentum.

I enjoy helping people move faster.

But somewhere along the way, I realised I had fallen into a trap many founders quietly experience.

I was trying to push people toward opportunities they weren’t even asking for.

I would see someone’s strengths clearly before they saw it themselves. I could often identify their strongest positioning, the direction with the highest potential leverage, or the opportunities sitting right in front of them.

Sometimes I helped structure their branding. Sometimes I opened doors. Sometimes I provided platforms, systems, tools, introductions, or guidance.

And yet, despite all of that support, very little happened.

Not because the opportunities weren’t real. Not because the systems were broken. But because ownership never fully materialised.

That was difficult for me to accept at first.

As founders, especially those who enjoy building communities or mentoring others, we often believe that if we provide enough support, enough tools, or enough encouragement, people will eventually move.

But eventually I realised something important: You cannot want success more than the other person does.

Also Read: Everyone wants AI in their product, but few know why (and when it actually works)

AI amplified this lesson for me

Ironically, AI made this reality even clearer.

Today, we live in a world where access has become incredibly democratised. People now have access to tools that previously required entire teams.

AI can help generate content, automate workflows, brainstorm ideas, accelerate execution, organise operations, and dramatically reduce friction.

In my own businesses, AI has significantly accelerated the speed at which I can move.

When I built earlier ventures years ago, reaching the first meaningful revenue milestones took months of experimentation, uncertainty, and manual effort.

Today, execution happens much faster.

Part of that comes from experience. Part of it comes from pattern recognition. And part of it comes from AI systems like Seraphina, my AI-powered digital twin, which helps me structure ideas, streamline workflows, and move from concept to execution far more efficiently.

But AI only accelerated the movement that already existed.

It did not create the movement itself.

That distinction is critical.

A trained AI is similar to a trained team. It amplifies direction, speed, and execution. But it still requires initiative, clarity, and decision-making from the person using it.

AI can reduce friction. It cannot manufacture discipline.

AI can accelerate execution. It cannot replace ownership.

And I think that is where many people misunderstand both entrepreneurship and AI today.

Buying tools is not the same as building. Joining communities is not the same as executing. Consuming information is not the same as moving.

The people who benefit the most from AI are usually those who were already willing to take action in the first place.

Builders move before they feel ready

One thing entrepreneurship taught me very early is that progress compounds.

Nobody starts at 10,000 users. Nobody starts fully prepared. Nobody starts with complete certainty.

You start with zero.

Then you learn. Then you adjust. Then you improve. Then you repeat.

Over time, the speed compounds because experience compounds.

That is why founders who have built before often move faster the next time around. The execution muscle becomes stronger. Pattern recognition improves. Decision-making sharpens.

But none of that happens without movement.

And that is why I eventually stopped trying to carry people who were unwilling to carry themselves.

Not because I stopped believing in people. Not because I stopped caring.

Also Read: The agentic shift: Why AI agents are rewriting the rules of ERP software in Singapore and Malaysia

But I finally understood the limits of what founders, mentors, AI systems, and communities can realistically do for someone else.

We can open doors. We can provide tools. We can shorten the learning curve.

But we cannot walk the path for them.

Leadership without self-sacrifice

I still believe deeply in helping people.

I still believe technology and AI can empower everyday individuals to build businesses, create freedom, and accelerate opportunities that previously felt inaccessible.

But I no longer believe it is my responsibility to save people from themselves.

That realisation changed how I lead, how I build communities, and how I approach growth.

Today, I focus less on convincing people to move and more on supporting the people already moving.

Because ownership changes everything.

And in an era where AI can help almost anyone execute faster than ever before, ownership may become the most valuable skill of all.

—

The post AI can accelerate execution, but it cannot replace ownership appeared first on e27.

Bitcoin’s retreat to US$76,632.16 reflects more than a routine correction. It captures a moment when geopolitical friction, macro uncertainty, and technical structure converged to test market conviction. The trigger came from escalating tensions between the United States and Iran. A social media warning from Donald Trump stating that time is running out for Tehran abruptly shifted sentiment.

Risk assets wobbled as Brent crude surged above US$112 per barrel before cooling toward US$107 to US$109, following diplomatic appeals from Saudi Arabia, Qatar, and the UAE that prompted a temporary pause in military action. That energy spike reignited inflation concerns and pushed expectations toward a higher-for-longer Federal Reserve policy, a headwind for any asset that thrives on abundant liquidity.

The macro shock exposed fragile positioning in crypto markets. Over US$607 million in bullish long positions were forcefully liquidated within 24 hours, part of a broader US$677 million wave of leveraged crypto long liquidations. When price fails to hold key levels, algorithmic selling and margin calls can accelerate moves far beyond fundamental justification. Bitcoin’s inability to clear its 200-day moving average near US$82,000 added technical pressure.

That rejection dragged the asset down to a critical support zone around US$76,000. Analysts note this level must hold to prevent a steeper structural breakdown toward US$65,000. The 200-week moving average near US$69,000 serves as a long-term trend reference, not a magnetic target price to be hit. Moving averages smooth past action; they do not dictate future paths.

The current weekly chart signals weakening momentum rather than outright capitulation. Price trades below shorter-term exponential moving averages but remains well above the 200-week trend line. The MACD indicator appears relatively controlled, suggesting the selloff lacks the extreme divergence often seen at major bottoms or tops. In strong trends, Bitcoin frequently establishes higher lows long before testing its slowest averages.

A move toward the low US$70,000s remains realistic if risk sentiment deteriorates further, but declaring US$61,000 inevitable simply because the 200-week moving average exists feels oversimplified. Markets respect context, and right now that context includes a regulatory landscape that is quietly evolving.

While traders navigate short-term volatility, Washington advanced a potentially transformative piece of legislation. The Digital Asset Market Clarity Act, known as the CLARITY Act, cleared a key hurdle when the Senate Banking Committee approved it in a bipartisan 15 to nine vote. This markup represents the first time a comprehensive crypto market structure bill has gained such momentum in the Senate.

The legislation aims to split oversight between the SEC and CFTC, define which digital assets qualify as digital commodities, and establish clearer registration and compliance frameworks for exchanges, brokers, and custodians. Provisions like a mature blockchain test and safe harbours for developers and noncustodial wallets seek to protect open source projects and peer-to-peer usage. If enacted broadly as described, large networks such as Bitcoin could receive clearer commodity treatment, easing institutional participation and exchange compliance.

Significant hurdles remain before the CLARITY Act becomes law. The bill must be merged with a separate Senate Agriculture Committee version, then secure 60 votes on the Senate floor, which requires at least seven Democratic votes. Ethics disputes over officials’ crypto holdings, the treatment of DeFi protocols and stablecoins, and a tight calendar window from June to early August, before recess and election politics intensify, all pose challenges.

Galaxy Digital’s research arm currently estimates a three-in-four chance that the bill becomes law in 2026, with an optimistic window for a presidential signature around early August if Congress moves quickly. For crypto participants, the critical signal will be whether Senate leaders schedule and win that 60-vote floor passage in the coming weeks. Without it, current momentum can still stall.

Global financial markets mirrored this fragmentation on 19 May 2026. US equity indices finished mixed as money rotated out of high-flying technology names and into defensive assets. The S&P 500 edged down 0.07 per cent to 7,403.05 while the Nasdaq Composite slipped 0.51 per cent to 26,090.73, dragged by a sharp correction in semiconductors. The Dow Jones Industrial Average gained 0.32 per cent to 49,686.12, supported by energy and traditional industrial components. Fixed-income markets drove much of the anxiety.

The US 10-year Treasury yield briefly breached 4.60 per cent, a fresh one-year high, while 30-year yields hovered above 5.10 per cent. Hotter-than-expected inflation metrics tied to Middle East tensions led traders to price in no 2026 rate cuts, with some shifting bets toward a potential hike later this year. International bond markets echoed the stress, with Japanese Government Bond 30-year yields touching multi-decade highs and UK Gilts experiencing similar spikes.

Sector performance highlighted the rotation. Memory chip and AI infrastructure names were hit hard after Seagate management expressed near-term supply-chain and demand constraints. Seagate fell roughly seven per cent to eight per cent, Micron declined six per cent, and Nvidia slipped two per cent ahead of its highly anticipated earnings release.

Meanwhile, defensive sectors and energy giants like Chevron gained ground, helping rescue the Dow. The equal-weighted S&P 500 notably outperformed its tech-heavy cap-weighted counterpart, underscoring the breadth of the rotation. In commodities, Brent crude cooled slightly as geopolitical fears eased marginally, while spot gold managed a slight rebound near US$4,589 per ounce, finding support from central bank accumulation despite a firmer US dollar.

These crosscurrents matter for Bitcoin’s path. The asset does not trade in isolation. It reacts to real yields, dollar strength, risk sentiment, and regulatory signals.

The post Oil spikes, bonds crash, Bitcoin drops: Here is what comes next appeared first on e27.

Southeast Asia’s digital payments narrative has always been too tidy. Consumers tap wallets, merchants display QR codes, and venture capital celebrates another fintech unicorn. But the IDC and 2C2P study exposes an uncomfortable reality: one-third of SMEs across the region, including in Singapore, still depend heavily on cash.

This is not a story about backward businesses resisting progress. It is about digital infrastructure that solves the checkout moment but ignores everything around it. Integration remains complex, fees are unpredictable, fraud concerns are legitimate, and settlement is often too slow for cash-flow-sensitive small businesses.

The Singapore data is particularly revealing. Here is a market with world-class infrastructure, high digital literacy, and mature fintech ecosystems; yet its SMEs report cash dependence levels matching Vietnam’s. That suggests the problem is not just about pipes and rails. It is about trust, workflows, and whether digital systems actually make running a business easier or just add another layer of operational friction.

For all the investment pouring into Southeast Asian fintech, the real opportunity may lie not in building another wallet, but in solving the messy back-office realities that keep merchants reaching for the cash drawer. Until digital payments fix the full stack, not just the front end, cash will remain the region’s most stubborn payment rail.

REGIONAL

SEA’s digital payments boom has a dirty secret: SMEs still run on cash: A new IDC and 2C2P study projects digital payments will reach 97% of SEA e-commerce by 2029, yet many SMEs remain cash-dependent, exposed by operational friction, security fears, and high fees holding back the region’s full payments transition.

Why SEA’s SMEs are falling out of love with bank-led payments: Banks remain the dominant payment provider for 79% of SEA SMEs, yet 88% are eyeing a switch, citing slow settlements, high fees, and weak integration — signals that legacy dominance increasingly reflects inertia rather than genuine loyalty.

Lightrock bets US$500M on energy access across SEA: London-based Lightrock’s Accelerate7 fund targets growth-stage clean energy companies, with Singapore’s TRIREC leading SEA deployment; 45M people in the region still lack electricity, and 250M rely on harmful cooking fuels, underscoring urgent demand.

JustCo eyes S$100M IPO on Singapore Exchange mainboard: The flexible workspace operator targets a market capitalisation of S$459.9M post-listing, issuing 32.1M shares at S$0.94 each, with cornerstone investors committing to 74.3M additional shares amid growing demand for hybrid work infrastructure across APAC.

Funding Societies and Boost Bank launch secured SME loans in Malaysia: The partnership enables Malaysian SMEs to borrow against industrial or residential properties for working capital and expansion, with Funding Societies originating financing supported by Boost Bank’s balance sheet, targeting asset-rich firms facing persistent credit gaps.

Singapore’s non-oil domestic exports surge 24.5% in April: Electronics shipments jumped 66.7% on AI-driven chip demand, while pharmaceuticals and machinery lifted non-electronics 10.9%; economists cautioned that rising energy and freight costs from the Iran conflict could slow momentum later in the year.

---

INTERVIEWS & FEATURES

PixVerse’s Jaden X: AI video’s biggest opportunity isn’t Hollywood: The co-founder of the 100M-user AI video unicorn says competitive advantage lies in combining a strong foundation model with deep productisation for everyday creators, not just professionals, with P2P sharing on messaging apps driving global growth across 177 countries.

Inside Inch Chua’s Myles: The AI boyfriend challenging love: Singaporean artist Inch Chua’s AI companion work raises urgent questions for founders about monetising emotional dependency; she argues AI companions optimise for retention over genuine human growth, warning that business models rewarding dependency will always undermine ethical intentions.

Venture debt in SEA: Non-dilutive capital with hidden legal strings: Growth-stage founders exploring venture debt must navigate security agreements granting lenders first claim on assets, warrant dilution, and restrictive covenants — legal obligations that can constrain strategic decisions for the full two-to-four-year loan term.

SEA consumers demand AI that connects, not just computes: SleekFlow’s whitepaper reveals 80% of SEA businesses deploy AI in customer service, yet 73% of shoppers prefer human-managed AI; Indonesia leads personalisation-driven purchase intent at 86%, while 45% of consumers expect responses within three minutes.

---

INTERNATIONAL

Meta moves 7,000 workers into AI roles ahead of job cuts: CEO Mark Zuckerberg is restructuring Meta into flatter, smaller teams focused on AI agents and apps, while cutting approximately 8,000 jobs on May 20, framing the layoffs as an efficiency drive to free up capital for accelerating AI investment.

Uber and Naver bid up to US$5.34B for South Korea’s Baemin: The 8-to-2 consortium is seeking full acquisition of Baedal Minjok from Germany’s Delivery Hero, with no final decision confirmed; the deal would mark one of the largest food-tech consolidations in Asia and signals Uber’s deepening regional platform ambitions.

Anthropic acquires Stainless in deal reported at over US$300M: The Claude-maker snapped up the Sequoia- and a16z-backed SDK tooling startup founded by a former Stripe engineer, while winding down hosted Stainless products; existing customers retain rights to their already-generated SDKs.

Bain Capital closes Asia Fund VI at US$10.5B, beating target: The firm surpassed its original US$7B target as it marks 20 years investing across Japan, India, China, Australia, and South Korea, with internal stakeholders serving as the fund’s largest investor group, signalling strong conviction in Asia’s continued growth trajectory.

Shein acquires Everlane in deal valuing the brand at US$100M: The fast-fashion giant bought the US apparel retailer from L Catterton to address roughly US$90M in debt, with common shareholders receiving no payout and preferred shareholder terms remaining unclear, as Shein pushes deeper into mainstream Western retail.

Revolut launches Dogecoin-themed debit card across UK and EU: The fintech’s new physical crypto card works wherever Visa and Mastercard are accepted with no extra exchange fees, entering a growing market alongside Coinbase and Crypto.com as Revolut simultaneously pursues banking licences in both the UK and US.

V-Green signs EV charging MOUs to expand in the Philippines: VinFast founder Pham Nhat Vuong’s V-Green plans 600 charging and 1,200 battery-swapping stations in Bataan province, while also partnering with Clean Fuel to install chargers at high-traffic fuel stations in Dasmariñas, Cavite.

---

CYBERSECURITY

CrowdStrike reports 43% rise in finance hacks globally: Financial sector intrusions surged over two years as 423 firms appeared on ransomware leak sites in 2025, up 27% year-on-year; North Korea-linked hackers alone stole US$2.02B in digital assets, increasingly leveraging AI-based deception to target institutions.

Borderless work, boundless risk: Securing the hybrid future in SEA: As the Philippines, Singapore, and Thailand embrace digital nomad policies, security must evolve beyond passwords to continuous device-level verification; Thailand’s data regulator imposed THB 21.5M in PDPA fines in 2025 alone, signalling the rising cost of compliance failure.

---

SEMICONDUCTOR

Tata Electronics and ASML partner for India chip fabrication: ASML will supply lithography tools, training, and supply chain support for Tata’s planned US$11B fab in Dholera, adding to existing partnerships with PSMC, Tokyo Electron, Rohm, Merck, and Intel as India accelerates its semiconductor ambitions.

Nvidia’s Jensen Huang expects China to reopen to US AI chips: Speaking after joining Trump’s Beijing delegation, Huang said H200 chip sales to China remain stalled despite US Commerce Department licences, as Beijing prioritises domestic semiconductor self-sufficiency and continues backing local firms including Huawei over foreign chip suppliers.

Hanmi Semiconductor to open US unit in San Jose by end of 2026: The South Korean equipment maker is establishing Hanmi USA to capture rising AI chip demand, with thermo-compression bonder orders concentrated in Q2 and HBM4 mass production expected to sustain momentum through the second half of the year.

---

AI

The US$2.5T bet: Why AI capital will mostly reward users, not builders: Gartner forecasts US$2.5T in global AI spending in 2026, yet historical infrastructure booms show value migrating to users, not builders; hyperscaler capex at 2.2% of US GDP, combined with rapidly depreciating GPU assets, suggests a capital cycle correction looming for infrastructure investors.

The shadow ledger: Why AI governance is now an enterprise revenue issue: With 82% of CIOs lacking verifiable oversight of AI agent actions and a 56.4% year-on-year rise in AI-related operational incidents, ungoverned deployments are generating hidden liabilities costing mid-market firms US$200K–US$400K annually in manual corrections.

Enterprise AI hits sovereign wall as data jurisdiction tightens: NTT DATA’s 2026 Global AI Report finds 95% of firms consider sovereign AI important, yet only 29% are acting on it concretely; 60% of AI leaders cite cross-border data restrictions as a major barrier, exposing a critical gap between ambition and architecture.

Alibaba and Tencent race to dominate AI-powered digital gateways: China’s tech giants are spending billions to make AI agents the primary interface for shopping, work, and communication, replacing search and super-app paradigms with conversational tools that interpret intent, recommend products, and complete transactions autonomously.

Singapore ranks in OpenAI’s top 5 markets for Codex adoption: OpenAI confirmed the city-state’s position based on weekly active users and blended token engagement, as the AI coding tool surpassed 4M global weekly users; OpenAI is also onboarding enterprise teams through partnerships with Accenture, Capgemini, and PwC.

Everyone wants AI in their product, but few know why it works: Most AI features are added due to market pressure rather than clear user need; genuine value emerges only when AI removes friction, accelerates key workflows, or improves decisions in ways users notice — not when it simply signals modernity to investors or competitors.

South Korea pilots program linking AI startups with trained talent: The Ministry of SMEs and Startups opened applications for a scheme offering up to US$134,000 in commercialisation funding to AI startups that hire government-trained graduates, targeting 80 firms across five deep-tech sectors amid a survey finding 57.3% of startups cite AI talent shortages as a major challenge.

Crypto and equities slide as geopolitical and macro pressures mount: Bitcoin fell 0.96% to US$77,388 while the S&P 500 dropped 1.24%, with US-Iran tensions pushing Brent crude past US$110 per barrel; the CLARITY Act’s Senate committee passage triggered US$980M in crypto liquidations as overleveraged traders unwound positions on the regulatory news.

Bitcoin rallied on regulation: Why the CLARITY Act changes everything: The Senate Banking Committee’s 15-9 approval of H.R. 3633 drove Bitcoin up 2.45% to US$81,511, resolving CFTC-SEC jurisdictional ambiguity and establishing stablecoin reward frameworks, with prediction markets assigning a 73% probability of full Senate passage.

---

THOUGHT LEADERSHIP

How earned media drives AI search visibility in ASEAN: Ahrefs’ analysis of 75,000 brands found web mentions matter three times more than backlinks for AI search visibility; with ASEAN’s AI market projected to reach US$80B by 2031, B2B brands must prioritise consistent, expert-led earned media placements to remain visible in AI-generated answers.

What is a brand and why it matters more than ever for startups: Grab’s reported brand value of US$1.1B against a US$12-15B market cap illustrates how intangible assets drive startup valuations; founders who build deliberate brand narratives from day one attract capital more easily and command stronger pricing power than those treating brand as a marketing afterthought.

The virtue of the closed door: Differentiation by intentional incompatibility: In an API-economy that rewards interoperability, true competitive defensibility comes from proprietary data structures, talent lock-in, and forced workflow divorces from competitors, turning switching costs into a strategic moat rather than building bridges that make replacement easy.

From seashells to tokens: Why 2026 could be the inflection point for money: With SWIFT launching a blockchain-based cross-border ledger, Malaysia piloting stablecoins, and APAC on-chain transaction value tripling to US$244B in 30 months, tokenisation is shifting from speculative momentum to a systemic convergence of infrastructure, regulation, banking, and sovereign adoption.

The post Ecosystem Roundup: Digital on the surface, cash underneath appeared first on e27.

Nearly half of businesses across the Asia-Pacific (APAC) region are directing significant capital toward Agentic AI systems, with new research showing that budgets for the tech are scaling faster than those for Generative AI at a comparable stage of its development.

According to a report by Omdia, the tech research and advisory arm of Informa, 42 per cent of organisations surveyed are allocating US$1 million or more to AI agents over the next 12 months. The findings, published in a special report titled Staying Ahead in the AI Era, were released ahead of the Asia Tech x Enterprise conference, scheduled to run from May 20 to 22 at Singapore EXPO.

The scale of investment reflects a broader shift in how enterprises across the region are approaching AI: moving away from exploratory pilots and towards systems capable of operating with minimal human intervention. Unlike conventional software, Agentic AI can initiate actions, coordinate workflows, and execute multi-step tasks autonomously, without requiring continuous human oversight at each stage.

Omdia’s research suggests this operational autonomy is fundamentally altering how businesses structure accountability. The report describes the shift as a break from a long-standing tech contract in which humans make decisions, and software supports them. As agentic systems take on more of the decision-making burden, organisations are under pressure to redefine roles — concentrating human contribution on exception handling, ethical judgment and systemic oversight rather than routine task management.

Also Read: The agentic shift: Why AI agents are rewriting the rules of ERP software in Singapore and Malaysia

The investment surge is also reshaping the physical infrastructure underpinning AI deployment. Traditional IT systems across Asia are increasingly being repurposed or replaced by what the report terms “AI Factories” — purpose-built facilities designed to produce intelligence at an industrial scale, continuously. This infrastructure buildout is running in parallel with advances in physical AI, as humanoid robotics moves into a more mature commercial phase in 2026. South Korea has made significant commitments to its robotics ecosystem, while Taiwan has positioned itself as a key contract manufacturer for global humanoid robotics vendors.

However, the pace of investment is not unlimited. Omdia found that 82 per cent of businesses surveyed are prepared to commit larger budgets, but only when measurable and defensible returns can be demonstrated. The findings point to an enterprise environment that is cautiously optimistic — willing to scale, but demanding clear evidence of business value before doing so.

Security considerations are also emerging as a parallel priority. As AI systems become more deeply embedded in enterprise operations, organisations face a four-part cybersecurity challenge spanning the use of AI to strengthen security, AI as a security tool, threats posed by adversarial AI, and the protection of AI systems themselves.

A further complication looms: researchers project that quantum computing could break widely used encryption standards, including RSA and AES, by 2030. Approximately 32 per cent of organisations surveyed are already exploring quantum-resistant protections in anticipation of that threshold.

Also Read: Agentic economy: The real promise of AI and crypto convergence

“The decisions enterprise leaders make in the next 18 months will define their competitive position,” said Joyce Wang, Event Director for Asia Tech x Singapore at Informa. “By mapping out these critical shifts — from agentic AI to quantum readiness and physical AI — we are ensuring that when delegates arrive at the Singapore EXPO this May, they are ready to transform these innovations into measurable business impact.”

The post Nearly half of APAC enterprises commit US$1M+ to agentic AI appeared first on e27.

Real estate is one of the largest industries in the world, and one of the least changed. It still runs on 1950s premises, sold as a “human-to-human, relationship-driven” business. In reality, agents in the technology era rarely deliver actual value to property transactions: they sit in the middle, gatekeeping information. Sitting in the middle of every transaction, humans have become the biggest bottleneck: the source of delays, ambiguity, and uncertainty in a process that should be fast, transparent, and end-to-end digital.

In Thailand, that friction has a direct price tag. For instance, landlords pay the equivalent of one to two months’ rent as commission for a successful introduction. Almost all of it flows to agent fees, marketing spend, and back-office overhead. This leaves brokerages with not over 10% gross margin ceiling regardless of market or volume.

Superagent’s bet: the agent itself is replaceable with AI, and the unit economics of doing so are dramatic.

The bet comes from someone who has spent 15+ years inside the system. Yuriy Braterskyy, Superagent’s founder and CEO, has held senior operating roles across Southeast Asia’s leading property marketplaces. Early on, he was COO of Hipflat (scaled over four years and sold to Dot Property). Then, he was Operations and Sales Director at Dot Property post-acquisition. Later, he became CEO of Seekster (acquired by True Digital, a major Thai corporation). Together, the platforms he ran used to help over a million people a month searching and improving their homes. He has concluded that for the majority of deals, the agent layer no longer needs to be there.

A new architecture for real estate

Superagent is the AI technology layer powering end-to-end real estate transactions, replacing human agents with AI across marketing, sales, negotiation, and closing. Superagent’s Bangkok rental platform is the first vertical built on this stack. The same technology deploys into new markets (Singapore, Indonesia, Malaysia, Australia and many more) as full AI-native real estate verticals.

AI handles the entire admin and operations layer end-to-end. This includes marketing, lead qualification, listing onboarding, omnichannel communication, matching, follow ups, scheduling, and negotiation. Humans step in only at the irreducibly physical last mile: touring properties and closing deals.

“Our bet is structural: AI runs the brokerage, humans only tour and close. Our big vision is property transactions happening directly between buyer and seller, with our AI as the ecosystem in between, removing friction the technology era should have eliminated long ago.”

– Yuriy Braterskyy, Founder & CEO, Superagent

Also read: Startups driving AI automation, fintech, and accessibility gather at Echelon Singapore 2026

Built AI-first

Y Combinator opened its 2026 Requests for Startups with a single sentence: “AI has stopped being a feature and started being the foundation.” YC’s frame for the new generation of companies is direct: “AI-native companies that don’t sell software—they sell the service.” Build the AI first; design the business model around what it can do.

Superagent followed exactly that arc. First as a SaaS, where the AI sales engine earned paying customers from real estate businesses across Thailand, Malaysia, Singapore, Australia, and the US. Then turned inward and now powering an end-to-end rental vertical in Bangkok, where the AI runs the entire operation with closed deals at software-grade margins.

Why now: A generation that rents

Renting is no longer a way station to ownership; for a growing share of the world, it is the destination. The IMF, surveying 40 countries over 50 years, calls the affordability deterioration of the past two years “sudden.” In Southeast Asia, the numbers are more extreme than anywhere else. Bangkok’s price-to-income ratio sits at 28.7×, Manila’s at 35.9×, against a standard affordability threshold of 5×. Over 66% of Thai Gen Z and Gen Y now prefer renting to buying, a cultural reversal that CBRE Thailand’s head of research describes as a shift from “ownership as stability” to “ownership as liability.” 73% of Singaporean Gen Z view homeownership as a longer-term goal, with affordability as the dominant obstacle. Two out of three Gen Z Indonesians are pessimistic about buying within three years; Jakarta rental demand surged 55% in a single quarter.

Affordability is only half the story. 83% of Gen Z renters say renting lets them save for life experiences. Specifically, flexibility, mobility, and remote work have rewritten what young adults want from a home. Location commitment is now a cost, not a benefit.

In Southeast Asia, the global trend lands on top of a regional accelerant. The global digital nomad population crossed 40 million in 2024, with extended stays of 30+ days up 42% over pre-pandemic levels and five of the world’s top ten remote-work cities sitting in the region. Superagent is built for the world after this shift, where transactions are frequent, customers are mobile, and the volume sits in rentals. AI is the only operating model that scales at the unit economics this volume demands.

The proof: Nearly 7× the industry’s unit economics

Superagent’s AI-native brokerage runs at a confirmed gross margin above 65%, aggregated across every property deal closed to date. The live business is Bangkok, where Superagent operates an end-to-end AI-run rental platform: multiple leases closed in the first month of operation, over US$5,000 in commission revenue within weeks of launch, its current pipeline scaling to multiple deals per week and accelerating.

The structural ceiling for human-agent brokerage margins is around 10%, regardless of market. PropNex Limited (SGX:OYY), Singapore’s largest residential agency by transaction volume (64.2% market share in 2024), reported a 9.1% gross margin for FY2024. APAC Realty Limited (SGX:CLN), parent of ERA Realty Network and the second-largest Singapore agency, reported 8.9%. These are the highest-quality regional brokerages with publicly audited financials in the region. Superagent runs at nearly 7× that ceiling, on every deal it closes.

Also read: AI, sustainability, and digital transformation leaders at Echelon Singapore 2026

A market that absorbs the bet

Southeast Asian residential real estate is a multi-trillion-dollar market across 700 million people in the world’s fastest-urbanizing region — and it runs almost entirely on legacy brokerage infrastructure. The same structural inefficiency that produces 9% gross margins in Singapore’s most sophisticated listed agencies exists in every major city across the region.

Thailand is the validation market. The Thailand residential real estate market is US$30.2 billion in 2025, projected to reach US$40.7 billion by 2031 (Mordor Intelligence). Bangkok represents 45.5% of it. Rental is the fastest-growing segment, expanding at a 5.88% CAGR, with prime rents up 6.5% YoY and gross rental yields holding above 6%. That is large enough to prove the model at scale. It is not the ceiling.

The SaaS phase already demonstrated the architecture travels: paying customers from real estate businesses in Thailand, Malaysia, Singapore, Australia, and the US ran on the same stack before Superagent turned it inward. The replication logic is straightforward: every market where agents sit between landlord and tenant, charging a month’s rent for a single introduction, is a market this model can enter.

Build by a small, technical team

Superagent has raised US$400,000 in pre-seed from lead investor Iterative, GenAI Fund, Innospace, and multiple angel investors from the real estate industry. Iterative’s Summer 2025 batch was one of the most competitive accelerator cohorts in Southeast Asia, with a ~1% acceptance rate. The team has also been selected for the AWS Spotlight programme and Google for Startups.

“Most AI systems for real estate today are a simple chatbot bolted onto a CRM. Superagent’s is an intelligent multi-agent system that owns the entire transaction state (supply, demand, matching, scheduling, handoffs) and executes its decisions across all customer touchpoints in over 40 languages, in real time. Existing established brokerages cannot retrofit this in their bloated operations.”

Ranjit Nagaraj, Founding AI Engineer, Superagent

What’s next

Having confirmed early product market fit, Superagent is now raising more capital to establish a dominant position in Bangkok ahead of an international seed. The capital deploys the same playbook market-by-market: the technology layer as a full AI-native vertical across countries in SEA and APAC.

The longer-term bet is bigger than market share. The endgame is that the majority of property transactions happen directly between seller and buyer, with Superagent’s AI as the ecosystem in between, removing the up-to-12%-of-sale-price commissions, or the full month’s rent charged for a single introduction, that the technology era should have eliminated long ago. This is non-zero-sum disruption: it grows the industry, rather than just redistributing what is already there.

Open conversations: Closing the round in June

The current pre-seed allocation is targeted to close before the end of June. First-call window: through the end of May 2026. Strategic investors, early stage VCs and angel investors are invited to engage now.

• Email Yuriy directly: yuriy@superagent.co
• Visit superagent.co
• Meet the team in person at their booth at Echelon Asia Summit, Singapore, 3-4 June 2026.

The region is evolving quickly, and Echelon 2026 offers the right place at the right moment to be part of what comes next. Register here to join the conversation.

—

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

The e27 team produced this article

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: Superagent

The post Superagent: The AI-native real estate brokerage running at 65%+ gross margin in an industry capped at 10% appeared first on e27.

Southeast Asia’s quick commerce market reached US$7.3 billion in gross merchandise value in 2025, accounting for 4.6 per cent of the region’s e-commerce market but less than 1 per cent of total retail, according to new research from Momentum Works.

The figures confirm what industry observers have long suspected: the infrastructure for rapid delivery exists, but consumer habits have yet to catch up.

Also Read: Why quick commerce is really about frequency, not speed

The report offers a corrective to the breathless narratives that have dominated quick commerce discourse in the region. Whilst platforms, retailers, and e-commerce players are converging on the same hyperlocal opportunity, the more critical question is not how big quick commerce will get, but how it will develop differently across Southeast Asia’s fragmented markets.

Unlike India, where platform-run dark stores have effectively leapfrogged offline retail to serve affluent metro consumers, or China, where a decade of food delivery investment built the hyperlocal rider density that underpinned quick commerce, Southeast Asia has dense existing retail networks that shape a fundamentally different growth path.

Grocery promised scale, but online penetration remains stuck at 4.2 per cent

Quick commerce in Southeast Asia was built on grocery. The category offered high purchase frequency, broad consumer relevance, and a clear value proposition: getting essentials delivered in under 30 minutes without leaving home. Yet online grocery penetration across the region stands at just 4.2 per cent, with Indonesia at 2.8 per cent and Singapore leading at 9.7 per cent.

That makes quick commerce a subset of an already-limited market. Consumers in Southeast Asia remain wedded to offline grocery shopping, whether through traditional wet markets, neighbourhood mini-marts, or modern supermarkets.

The challenge is structural. Offline retail in Southeast Asia works remarkably well for most consumers—mini-marts like Alfamart and Indomaret blanket Indonesian cities. Thailand’s convenience ecosystem is tightly controlled by conglomerates with little incentive to disrupt their own offline networks. Singapore has high modern trade penetration, but limited geographic scale.

In this environment, quick commerce cannot simply replace offline retail. Instead, it must extend it, layering on-demand fulfilment over infrastructure that already serves consumers efficiently. Platforms are now responding by expanding aggressively beyond grocery into general merchandise, personal care, pharmacy, and other categories.

E-commerce platforms double down: 3.4 per cent of GMV now quick commerce

Quick commerce is emerging as the next battleground for Southeast Asia’s dominant e-commerce platforms. According to Momentum Works, 3.4 per cent of the region’s e-commerce GMV is now fulfilled through quick commerce networks, a meaningful early indicator of where platform competition is heading.

Also Read: The future of social and quick commerce for developing countries

Shopee is leveraging ShopeeFeed’s fulfilment infrastructure to roll out instant delivery across multiple markets. Lazada has launched on-demand grocery fulfilment through RedMart Now in Singapore. Grab is expanding categories offered on GrabMart from grocery to general merchandise and beauty by working with supermarkets and retailers.

The strategic shift is clear: platforms that built their businesses on next-day or same-day delivery are now racing to offer sub-hour fulfilment. What distinguishes Southeast Asia from other markets is that quick commerce here is built on existing infrastructure rather than created from scratch. E-commerce platforms are not building vast networks of dark stores. Instead, they are integrating with online grocery operations, supermarkets, mini-marts, and retailers, creating a more distributed, asset-light model.

This approach is operationally complex but potentially more sustainable. It avoids the heavy capital expenditure of building and stocking warehouses, instead leveraging existing retail inventory and locations. The challenge is that integration and coordination are harder to execute than in a vertically controlled dark-store network.

Southeast Asia diverges from China and India’s playbook

Quick commerce accounts for 4.6 per cent of Southeast Asia’s e-commerce market, compared to 16.6 per cent in India and 7.4 per cent in China. That gap reflects different retail structures rather than simply a lag in maturity.

In China, Meituan and other platforms invested heavily in food delivery for over a decade, building dense hyperlocal rider networks that could be repurposed for grocery and merchandise delivery. In India, low organised retail penetration enabled platform-run dark stores to serve as modern retail for affluent urban consumers effectively.

Southeast Asia has neither of these conditions. Offline retail is already well-developed and fragmented. Food delivery networks exist, but are less dense than China’s. The result is a market where quick commerce must prove its value against alternatives that already work reasonably well.

Six markets, six playbooks: No regional model will scale

One of the report’s sharpest insights is that Southeast Asia’s quick commerce opportunity is fundamentally local, not regional. Each of the six major markets — Indonesia, Thailand, Singapore, Vietnam, the Philippines, and Malaysia — has distinct structural realities that shape each operator’s playbook.

In Indonesia, growth is more likely to come from e-commerce platforms than from mini-mart chains. Alfamart and Indomaret have limited incentive to expand quick commerce themselves, given the strength of their offline networks.

Thailand’s mainstream convenience layer is locked up by conglomerates, pushing the quick commerce opportunity toward vertical plays in specific categories rather than broad horizontal platforms.

Singapore meets the structural conditions for quick commerce — high income, dense urban geography, digitally savvy consumers — but is constrained by its market size.

Vietnam’s modern retail has built physical scale, but fresh-grocery habits remain anchored in traditional channels. The Philippines has the lowest modern trade penetration in Southeast Asia, with organised retail concentrated in malls. Malaysia’s retail is structurally fragmented, with no single dominant player.

The implication for operators is clear: a single regional playbook will not work. Success requires deep local knowledge, market-specific partnerships, and tailored strategies that account for each market’s unique retail structure, consumer behaviour, and competitive dynamics.

Demand, not supply, is the binding constraint

Perhaps the report’s most important conclusion is that Southeast Asia has the infrastructure for quick commerce (riders, stores, and platforms) but lacks the consumer habit. Fulfilment and supply are not the problems. Demand density is.

Also Read: SEA e-commerce surges to US$185B as video commerce becomes the new growth engine

Closing this gap requires sustained capital to fund price parity with offline and e-commerce. Speed is quick commerce’s natural value proposition, but in a structurally price-sensitive region, mass-market adoption requires pricing at or near parity with alternatives. That takes subsidies, which only well-capitalised platforms can afford.

The question is whether platforms will maintain the discipline to invest in demand generation without burning capital on unsustainable unit economics. Quick commerce in Southeast Asia is real, meaningful, and growing. But it remains in early stages, with the hard work of changing consumer behaviour still ahead.

The post SEA’s US$7.3B quick commerce market is solving the wrong problem appeared first on e27.