For over a decade, Sony Acceleration Platform has quietly supported more than 1,000 business development cases through Sony’s internal innovation arm. Its newest initiative, the Boundary Spanning Service, extends that experience across borders, connecting Japanese corporations seeking practical innovation partners with Singapore’s fast-moving startup ecosystem.

Developed in collaboration with e27, Boundary Spanning Service reflects a broader shift in how Japanese enterprises are approaching open innovation: not as a scouting exercise, but as a long-term, trust-based collaboration. e27 spoke with Sony Acceleration Platform team about what is driving this initiative, and what it means for founders on both sides of the corridor.

e27 sat down with the Sony Acceleration Platform leadership to discuss the vision behind Boundary Spanning Service and what it means for Singapore founders.

The vision and context

Could you share a little about what led to the creation of the Boundary Spanning Service – what need or opportunity did Sony Acceleration Platform observe that this initiative is designed to address?

The creation of the Boundary Spanning Service was driven by a strategic need to address a critical structural issue in business development: the tendency for promising projects to ‘run out of time’ before they can fully scale.

Drawing on over 12 years of experience supporting more than 1,000 business development cases within the Sony Acceleration Platform, we observed that one of the reasons why new businesses often fail is not lack of viability, but due to rigid corporate constraints.

To bridge this gap, Boundary Spanning Service was established to act as a ‘Boundary Spanner’. Sony’s own history is defined by absorbing external knowledge to drive growth; thus, facilitating open innovation is a natural extension of our DNA. Through the Boundary Spanning Service, we would like to provide a structured framework that connects Japanese corporations facing specific innovation bottlenecks with Singapore’s agile and highly capable startup ecosystem. By crossing organisational and geographic borders, Boundary Spanning Service lowers survival costs and accelerates real-world deployment, ensuring valuable business developments are not prematurely terminated.

From Sony Acceleration Platform’s perspective, what makes Singapore a meaningful starting point for this initiative? We’d love to understand what Sony Acceleration Platform sees in the Singapore ecosystem that feels relevant to Japanese corporates.

Singapore is uniquely positioned as our starting point because of its exceptional alignment with the strategic needs of Japanese enterprises. The Singapore ecosystem offers a rare combination of advanced technological capability, a highly robust and predictable regulatory environment, and seamless English-language business operations.

For Japanese corporations looking to mitigate risk and operational uncertainty in cross-border ventures, these factors provide an incredibly stable and efficient platform. Singapore’s startup community is not only highly innovative but also operates with global compliance and agility, making it the ideal counterpart for Japanese enterprises seeking reliable, fast-paced co-creation.

Also Read: e27 expands AI-powered business matchmaking with Sony Acceleration Platform collaboration

When Japanese corporations come to the Boundary Spanning Service, what kinds of challenges or aspirations are they typically bringing with them? Are there particular themes or sectors you’ve seen emerge?

Japanese corporations are not coming to Boundary Spanning Service for abstract technology scouting or passive trend-watching. Instead, they bring highly defined, practical operational bottlenecks and specific innovation challenges. They are actively seeking practical, market-ready solutions that can be integrated into their existing value chains.

While the technical fields vary, the underlying theme is the need for rapid digital transformation, advanced automation, and niche technological capabilities that can be fast-tracked for real-world, commercial deployment. They come with a genuine urgency to solve immediate business bottlenecks by leveraging the agility of external partners.

The nature of Japan-Singapore collaboration

How would you describe the spirit of collaboration that Boundary Spanning Service is designed to enable? What does a meaningful, productive engagement between a Japanese corporate and a Singapore startup tend to look like in practice?

Boundary Spanning Service is designed to move far away from low-value, one-off “transactional procurement”. A truly productive engagement must be grounded in mutual trust and reciprocal value creation.

In practice, this means establishing a collaborative framework where both parties act as equal partners. The Japanese corporate provides deep operational resources, industry expertise, and market access, while the Singapore startup provides the agility, speed, and disruptive technology needed to overcome the bottleneck. We foster an environment where cultural and organisational differences are structurally bridged, transforming potential friction into collaborative synergy.

Could you share a sense of what success looks like for both the Japanese corporate and the Singapore startup that comes through the Boundary Spanning Service? Even in broad terms, what outcomes feel meaningful to Sony Acceleration Platform?

Meaningful success is achieved when the collaboration translates into a viable, long-term business outcome. For the Singapore startup, this means successfully scaling their operations and entering the Japanese market backed by the massive distribution networks and credibility of a major corporate partner. For the Japanese corporate, it means successfully resolving a critical business bottleneck while absorbing the entrepreneurial agility and speed of the startup.

For Singapore founders

For a Singapore startup founder who is curious about Boundary Spanning Service, what qualities or characteristics tend to make for a strong and rewarding collaboration with Japanese corporate entities? What do you find matters most?

The most critical asset a founder can bring is a commitment to mutual alignment and long-term planning. Japanese corporate partners place an immense premium on quality standards, operational stability, and meticulous planning.

Instead of viewing these requirements as bureaucratic delays, successful founders recognise them as the very foundation required to achieve sustainable, enterprise-grade scalability. A willingness to understand these operational values, combined with transparent communication and professional patience, is what truly secures a rewarding, high-yield partnership.

Also Read: Global expansion is no longer about reducing information costs, it’s about reducing trust costs

What would a founder’s journey through the Boundary Spanning Service look like from start to finish – from the initial application through to the first conversation with a Japanese corporate? What should they expect in terms of timing and engagement?

We have structured the journey to be as seamless and high-probability as possible. Through our joint effort with e27, startups can apply through a highly streamlined, low-friction application process.

From late September, Sony Acceleration Platform will start sharing and recommending the startup’s profile to the participating Japanese companies who are interested in connecting with Singapore companies.

Is there anything you’d like to say to Singapore founders who may be open to this kind of collaboration but are perhaps unfamiliar with how Japanese corporates typically work or what they value in a working relationship?

It is important to understand that the thoroughness of Japanese corporate decision-making—which often involves extensive internal consensus-building—is a structural characteristic, not a personal hurdle.

While this front-loaded alignment process takes time and structured engagement, the payoff is unparalleled. Once a Japanese corporate commits to a partner and establishes mutual trust, that relationship translates into exceptionally stable, deeply committed, and highly scalable long-term support. We encourage founders to focus on building a robust foundation of quality and trust from day one.

Looking ahead

Looking ahead, what does Sony Acceleration Platform hope the relationship between Japanese corporate innovation and Singapore’s startup ecosystem will look like over time? What would feel meaningful to you personally?

Our ultimate vision is to see cross-border co-creation transition from a “special project” into a standard, daily operational model. Personally, it would be deeply rewarding to see this corridor become the default pathway for global-scale business development.

We highly encourage forward-thinking founders to leverage the Boundary Spanning Service to position themselves at the very forefront of this evolving bilateral corridor.

—

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

The post Bridging boundaries: Sony’s blueprint for Japan-Singapore innovation collaboration appeared first on e27.

Southeast Asia’s startup story is usually told through ride-hailing, fintech, and e-commerce. But a quieter, stranger, and potentially more consequential sector is taking shape in the region’s labs, farms, and hospitals: biotech.

Founders are using fermentation to turn waste into biomaterials, growing seafood and fat from cells, building AI tools for cancer care and heart scans, engineering crops for tougher climates, and developing diagnostics designed specifically for Asian populations. It is not an easy sector to build in; biotech demands patient capital, deep technical talent, complex regulatory navigation, and long commercialisation cycles.

Also Read: Asia’s biotech boom: Innovation, investment, and a new era of discovery

Yet the sheer breadth of companies now emerging from Singapore, Malaysia, Indonesia, Thailand, Vietnam, and the Philippines suggests the region’s life-sciences ecosystem has moved well past the experimental stage, and is beginning to challenge assumptions about where deeptech gets built.

Below is a list of 27 biotech firms that are redrawing the region innovation map:

RWDC Industries (Singapore)

RWDC ferments used cooking oil into PHA, a fully biodegradable biopolymer that’s meant to replace single-use plastic in straws, cutlery and packaging.

Protenga (Singapore/Malaysia)

Protenga runs “Smart Insect Farms” that turn organic waste into black soldier fly protein for aquaculture, animal feed and pet food.

Engine Biosciences (Singapore/US)

Engine Biosciences combines AI with wet-lab biology to map gene interactions and speed up cancer drug discovery.

Us2.ai (Singapore)

Us2.ai uses AI to automate the reading of echocardiograms, cutting a process that takes cardiologists many minutes down to under two.
Founders: James Hare, Dr Carolyn Lam, Dr Yoran Hummel and Paul Seekings

AMILI (Singapore)

AMILI runs Southeast Asia’s first gut microbiome bank, building an Asia-specific database to power diagnostics and personalised nutrition.

KYAN Technologies (Singapore)

KYAN applies “small data AI” to match cancer patients with the most effective drug-dose combinations, developed with NUS and UCLA.

ImpacFat (Singapore)

ImpacFat cultivates omega-3-rich fish fat from stem cells, aimed at alt-meat, cosmetics and supplements.

Qarbotech (Malaysia)

Qarbotech makes QarboGrow, a carbon-quantum-dot photosynthesis enhancer that boosts crop yields without genetic modification.

NLYTech Biotech (Malaysia)

NLYTech develops biodegradable, plastic-free packaging materials made from natural ingredients as an alternative to single-use plastics.

Vulcan Augmetics (Vietnam)

Vulcan builds affordable, modular robotic prosthetics designed to click together like building blocks.

Teora (Singapore)

Teora develops biologics that manage disease in agriculture and aquaculture without relying on chemical pesticides.

KINNVA (Singapore)

KINNVA is a synthetic-biology company using fermentation to turn waste streams into biochemicals for food, feed and cosmetics.

Sinhke (Vietnam)

Sinhke builds AI-powered hardware that measures shrimp larvae health before farmers commit to large-scale cultivation.

Virdalis (Singapore)

Virdalis is building a cultivation and data platform around duckweed, the world’s fastest-growing flowering plant, as a soy alternative for animal feed protein.

Allozymes (Singapore)

Allozymes runs an ultra-high-throughput microfluidics platform that screens millions of enzyme variants a day, effectively an “enzyme discovery engine” for pharma, food and chemical industries.

FathomX (Singapore)

FathomX is an NUS/NUHS spin-off building AI to improve the accuracy of mammograms, particularly for dense breast tissue common among Asian women.

REVIVO BioSystems (Singapore)

An A*STAR spin-off building “organ-on-a-chip” 4D human skin models to replace animal testing for cosmetics and pharma compounds.

Advanx Health (Malaysia)

Malaysia’s first consumer DNA-testing company, offering genetic reports on health risk, nutrition and fitness traits.

KosmodeHealth (Singapore)

An NUS Food Science spin-off extracting proteins and fibres from food-processing waste to formulate functional foods and biomedical bio-ink.

Imagene Labs (Singapore)

A subsidiary of Asia Genomics offering saliva-based DNA testing for personalised nutrition, skincare and fitness products across Asia.

Meatiply (Singapore)

Meatiply is a multi-cell-type cultivated meat company that produced Asia’s first cultivated smoked duck breast.

Singrow (Singapore)

An agri-biotech firm that cross-bred and gene-edited the world’s first tropical-climate strawberry, grown in an indoor vertical farm.

Dendrotonics (Philippines)

Develops biodiversity-restoration technology to make degraded land ecologically and commercially productive again.

QuikPath (Singapore)

Built a self-administered RT-PCR Covid-19 screening technology designed to scale rapid, accurate infection monitoring.

ETBio (Singapore)

Harnesses microalgae to build next-generation air filtration solutions.

Ternion Biosciences (Singapore)

Provides high-throughput cardiac safety screening assays used in preclinical drug development.

CloudSeq (Singapore)

CloudSeq runs an NRF-backed cloud platform built to handle big data for healthcare and agricultural genomics.

The post The 27 SEA biotech firms betting on cells, fermentation, and code appeared first on e27.

Something strange is happening in Vietnam’s engineering hiring market right now. Companies report talent shortages. Candidates report rejection rates higher than ever. Both are telling the truth.

The gap between them isn’t a supply problem. It’s a definition problem, a fundamental mismatch between what companies say they need, what their hiring processes actually select for, and what the engineers who will matter most in the next three years actually look like.

I’ve spent a decade placing engineers across Southeast Asia. I’ve seen this kind of misalignment before. I haven’t seen it this structural.

The shift nobody fully prepared for

Eighteen months ago, the hottest debate in engineering circles was whether “Vibe Coding”, the practice of generating code entirely through natural language prompts to AI, was a legitimate workflow or a shortcut for junior developers. That debate is now over.

It turns out both sides were right, and both sides missed the point.

Yes, Vibe Coding accelerated output. A GitClear analysis of over 211 million lines of changed code found that AI-assisted workflows boosted boilerplate writing by 25–50 per cent. Yes, it also introduced a technical debt crisis: refactoring rates dropped from 25 per cent in 2021 to under 10 per cent in 2024, while duplicated code quadrupled. Over 40 per cent of junior developers admitted to deploying AI-generated code they hadn’t read.

By early 2026, Andrej Karpathy, one of the people who popularised the Vibe Coding concept, had already moved on. He began describing what comes next: Agentic Engineering. Not writing code, not prompting AI to write code, but orchestrating autonomous AI agents: setting specifications, auditing outputs, managing feedback loops, and owning architectural decisions.

The role of a software engineer is shifting from executor to decision-maker.

That shift has a direct consequence for anyone responsible for hiring.

The talent gap is real, but it’s not the gap most people think

Vietnam’s numbers are striking. The country now has 530,000–600,000 software engineers in a workforce of over 1.2 million ICT professionals. Universities produce 55,000–60,000 IT graduates per year. On paper, this looks like abundance.

In practice, demand for new technology positions exceeds 500,000 roles annually, a structural mismatch approaching 10x. And that gap is widening, not closing, for two compounding reasons.

Also Read: Vietnam’s healthtech boom has a talent problem nobody is talking about

• First, traditional industries entered the race. Banking, retail, and manufacturing have accelerated digital transformation, competing directly with tech companies for the same engineering talent. The pool didn’t grow; the number of teams fishing in it multiplied.
• Second, semiconductors arrived. Both Ho Chi Minh City and Hanoi are now running significant semiconductor and chip-design initiatives; HCMC alone is targeting 3,000 specialist engineers, backed by programmes at ĐHQG and SHTP. Vietnam’s national target is 50,000 semiconductor engineers at university level by 2030, supported by government scholarships worth 1,300 billion VND (US$49.4 million) annually for approximately 30,000 learners.

The consequence? The best STEM graduates are no longer choosing software development as their default. The pool of candidates that tech companies relied on for the past decade is being redirected upstream.

The result: average time-to-fill for a Senior offshore engineer has stretched to 45–60 days. Retention for engineers over two years has fallen from 78 per cent in 2023 to roughly 65 per cent in 2026. Senior IT salaries have risen 50–70 per cent compared to 2024, with significant variation depending on specialisation and language ability.

The two mistakes companies are making right now

Mistake one: Hiring for the old role

Most job descriptions I see in 2026 are still optimised to find Task-Based Coders, engineers who execute well-defined tickets, follow established patterns, and stay in their lane. The interview process tests syntax, algorithms, and framework knowledge.

But the engineers who will deliver the most value in an Agentic Engineering environment are evaluated on completely different dimensions: system design judgment, the ability to audit AI-generated outputs, risk assessment, and the capacity to make independent technical decisions under pressure. These skills don’t show up on a LeetCode score.

The irony is that the very engineers companies need most are often screening out of traditional hiring pipelines, because they’ve spent recent years developing meta-skills rather than memorising framework internals.

Mistake two: Trusting language credentials over language capability

Vietnam’s tech talent market has a well-documented phenomenon I call the Paper Certificate Trap.

Language certifications, TOEIC 850+, JLPT N2, TOPIK 5, are treated as proxies for communication ability. In practice, they measure test-taking performance under controlled conditions. I have interviewed engineers with near-perfect TOEIC scores who go completely silent the moment a client asks a follow-up question in a technical meeting.

This matters because language ability is one of the strongest economic multipliers in Vietnam’s engineering market. Engineers with professional English (B2–C1), Japanese (N3–N1), or Korean (TOPIK 4–6) command salaries 30–50 per cent higher than peers with equivalent technical experience but limited to Vietnamese. A Senior AI/ML engineer with strong English can realistically earn US$3,800–US$6,000+ per month, a meaningful difference driven entirely by the ability to negotiate architecture directly with international clients.

Companies that can’t reliably identify genuine bilingual capability are paying a premium for a credential that doesn’t reflect reality, while missing engineers who have real cross-cultural communication skills but modest exam scores.

Also Read: Great talent is what happens after AI creates the first draft

What actually works

Move to skills-based sourcing

Replace credential screening with competency screening. Define the actual decisions and judgment calls the role requires, then design your process to surface those directly.

For senior roles in an Agentic Engineering environment, the relevant competencies are: Can this person write a system specification and defend it? Can they review a diff they didn’t write and identify the architectural implications? Can they set up a feedback loop between AI agents and quality gates?

None of these appears on a CV. All of them can be assessed in a structured technical conversation.

Implement live communication audits

For any role requiring cross-timezone collaboration or direct client contact, add a real-time communication component early in your process, not a written English test, but an actual technical conversation under mild pressure.

A 20-minute session where a candidate explains a system they’ve built, fields two or three unexpected questions, and works through a hypothetical trade-off out loud will reveal more than any certification score. Done well, this eliminates the majority of candidates who present strong paper credentials but lack genuine communication fluency, before you’ve invested weeks in technical rounds.

Match your hiring model to your actual risk profile

Not all talent gaps require the same solution, and the mid-2026 environment punishes generic approaches.

Early-stage teams prioritise flexibility over headcount permanence; access to senior expertise without long-term fixed cost is often more valuable than a full-time hire at a moment when product direction is still shifting. Growth-stage companies typically benefit from a hybrid structure: a stable core for culture and IP continuity, with flexible capacity to absorb demand spikes. Vietnam’s tech market has consistent biannual attrition cycles, with July historically the highest-churn month as mid-year reviews conclude and bonuses are paid, predictable volatility that hiring plans rarely account for.

Larger enterprises and foreign-invested companies face a different constraint: the gap between standing up a dedicated engineering function and actually integrating it. Whatever structure is chosen, the critical implementation principle is the same. Cultural and operational integration (shared tooling, CI/CD pipelines, daily standups) must begin from Day 1, not at the point of handover. Teams that delay this until a later phase consistently experience attrition at precisely the moment continuity matters most.

Also Read: Great talent is what happens after AI creates the first draft

The bigger picture

The scarcity that talent leaders are experiencing in 2026 is not a temporary supply shortage. It reflects a structural reclassification of what engineering capability means and a transition period where the market hasn’t yet developed reliable signals for identifying the new kind of engineer.

The companies that hire well in this environment will be the ones that invest in building those signals themselves: clearer definitions of what “decision-ready” means for their specific context, better processes for detecting genuine bilingual capability, and hiring models flexible enough to absorb the volatility of a market where the best people have more options than ever.

The companies that don’t will spend the next 18 months paying premium salaries for engineers who looked right on paper, watching their technical debt compound quietly in the background.

I’ve seen both outcomes. The difference is almost always made before the offer is signed.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

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

The post Vietnam’s tech talent market is broken and most companies are still hiring the wrong way appeared first on e27.

Thai founders sometimes ask why a Vietnamese startup with a comparable product or level of traction can appear to raise a larger funding round.

The answer is rarely that one startup is inherently better than the other.

Venture capital reflects the company being financed, but it also reflects the market surrounding it. Investors consider the size and growth of the domestic economy, the availability of follow-on capital and the likelihood of eventually selling their shares.

On these measures, Vietnam currently benefits from a stronger growth narrative.

This does not mean Vietnamese startups always raise more than Thai companies. Southeast Asian funding data remain incomplete, many transactions are undisclosed, and a few large deals can distort national totals.

But a broader distinction is visible: investors are often more willing to finance Vietnamese startups against expected growth. Thai founders are more frequently required to demonstrate regional scale before receiving comparable backing.

Investors price future growth

Venture capital is a wager on what a company could become several years from now.

That makes national economic expectations important, even when investors are evaluating an individual startup.

Vietnam’s economy expanded by 8 per cent in 2025, while the World Bank expects growth of 6.8 per cent in 2026. Thailand, by comparison, is expected to grow by about 1.6 per cent in 2026.

Economic growth does not determine whether a particular software, healthcare or logistics startup will succeed. But it affects the assumptions investors place around that company.

Vietnam offers a population of more than 100 million, rising household incomes, manufacturing expansion and growing demand for digital services. An investor can reasonably expect some companies to expand alongside the economy.

Thailand is wealthier and has stronger infrastructure in many areas. It is also home to sophisticated banks, retailers, telecommunications groups and industrial companies.

These are valuable assets for startups seeking customers and partnerships. But they can also make the venture case more difficult.

A Thai startup may need to displace established companies in a relatively mature market. A Vietnamese company may be able to grow by serving demand that is still being created.

As a result, the Vietnamese startup can sometimes receive more credit for future scale, even when the Thai company has stronger revenue today.

Also Read: Inside SEA’s AI gold rush: The 20 investors writing the biggest cheques

Market size changes the fundraising conversation

Vietnam’s population is significantly larger than Thailand’s. This gives consumer-facing companies a broader domestic market from which to build.

A Vietnamese startup can often present domestic expansion as a venture-scale opportunity. A Thai startup in the same category may be asked almost immediately about Indonesia, Vietnam, Malaysia or the Philippines.

Thailand’s market can produce substantial companies. But venture funds are not simply looking for good businesses. They need a small number of investments to generate unusually large returns across a portfolio in which many companies will fail.

This pushes investors towards businesses that can reach large markets.

For Thai founders, the result is an execution discount. Investors may believe that the domestic business is sound while assigning limited value to regional growth that has not yet been demonstrated.

This is why the first customer outside Thailand can matter so much. It shows that the company’s opportunity is not restricted by the size or maturity of its home market.

Capital follows other capital

The composition of the investor ecosystem also influences funding rounds.

Vietnam attracted nearly 150 active venture investors in 2024, according to the Vietnam Innovation and Private Capital Report. Funds from Singapore and Japan were among the most active international participants.

Funding remains difficult. Vietnamese technology startups experienced a sharp decline in investment after the global venture boom, and national private-capital figures often include large buyouts that are unrelated to early-stage startups.

The important point is not that Vietnam has unlimited capital. It is that a growing number of regional investors already include the country in their investment strategies.

Also Read: Inside Singapore’s startup boom: The 21 firms investors can’t stop funding

Venture capital depends on networks.

A seed investor wants to know who might lead the next round. A Series A investor considers whether growth funds will be available later. Every investor eventually asks who might acquire the company or purchase its shares.

When many funds already follow a market, investors know the potential co-investors, corporate buyers and later-stage funders. This makes rounds easier to assemble.

Thailand does not lack capital. It has independent funds, family offices, government programmes and a substantial corporate venture sector.

Large Thai companies can provide startups with distribution, customers, regulatory knowledge and technical expertise. Yet corporate venture capital is not always a substitute for independent institutional funding.

Corporate investors may prioritise strategic alignment over financial returns. They may avoid companies that compete with another group subsidiary or require several layers of internal approval before investing.

They may also be willing to join a round without leading it.

A lead investor sets the terms, conducts extensive due diligence and gives other investors confidence to participate. Without one, a startup may receive interest from several organisations but still fail to close a substantial round.

The shortage of investors able and willing to lead larger early-stage rounds remains one of Thailand’s most important financing constraints.

The exit question begins early

Founders often discuss exits as a distant issue. Investors consider them before making the first investment.

A venture fund earns its return when it can sell its shares through an acquisition, a secondary transaction or a public listing.

Thailand has a large stock exchange and some of Southeast Asia’s most powerful corporate groups. Yet the country has not developed a predictable exit path for venture-backed technology companies.

This can create a cycle.

Limited exits attract smaller funds. Smaller funds write smaller cheques. Startups then have less capital to expand regionally, making large exits even less likely.

Vietnam’s exit market is not mature either. Its improving public-market narrative does not yet provide a reliable listing route for technology startups.

However, Vietnam’s role in regional manufacturing, trade and supply chains gives strategic investors several reasons to acquire local technology, logistics and enterprise businesses.

Thailand has similar strengths in tourism, healthcare, food, energy, automotive manufacturing and services. The challenge is to connect these sectors to regional buyers rather than treating acquisition by a domestic conglomerate as the only possible outcome.

What Thai founders can control

Founders cannot change Thailand’s demographics, economic growth or fund structure. They can change how dependent their company appears to be on the domestic market.

Regional expansion must be presented as an operating plan, not a collection of flags in a pitch deck.

A Thai software company might follow an existing corporate client into Malaysia. A hospitality platform could expand through Thai hotel groups operating abroad. A healthcare startup could target countries with similar private hospital systems.

Internationally comparable metrics are also essential. Recurring revenue, retention, gross margin, customer acquisition costs and contribution margin help investors compare the company with businesses in other markets.

Thai founders should also approach regional investors before they urgently need capital. A fund that has followed a company for a year can evaluate its progress more confidently than one receiving a pitch shortly before the runway expires.

Most importantly, founders need to identify which investors can actually lead a round. Interest from corporate funds and smaller investors is useful, but it may not be enough to establish the valuation and bring the full syndicate together.

An expectations premium versus an execution discount

The difference between the two ecosystems is not that Vietnamese founders consistently build better companies.

Vietnam benefits from an expectations premium. Investors see a large market, faster economic growth and a growing network of international funds. They are sometimes willing to finance the scale a company may eventually achieve.

Thailand faces an execution discount. Startups are more often expected to show regional revenue, efficient economics and clear evidence that they can grow beyond the domestic market.

Both perceptions are incomplete. Vietnam remains exposed to trade disruption, regulatory risk and limited exits. Thailand has sophisticated infrastructure, strong corporations and real competitive advantages.

But investor narratives affect how capital is allocated.

Vietnamese startups can sometimes raise against the future investors expect their market to create. Thai founders are more often required to begin building that future before investors will pay for it.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

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

The post Why investors often back Vietnamese startups more aggressively than Thai peers appeared first on e27.

Southeast Asia is in the midst of a workforce transformation paradox that has quietly become the region’s most pressing business challenge.

Governments have invested billions in upskilling initiatives. Singapore alone has trained over 555,000 workers through SkillsFuture programmes. Across the region—from Indonesia’s digital transformation drive to Vietnam’s emerging tech ecosystem—organisations are spending heavily on employee development. SMEs are sending their teams to AI courses, data science bootcamps, and digital literacy programmes. On paper, the workforce has never been better prepared.

Yet when these trained employees return to their jobs, something breaks.

The CEO who approved the training gets a report that the new “AI-capable” team member isn’t delivering AI-ready outputs. The employee who completed certification feels anxious despite their credential. The hiring manager who reviewed a resume with “AI Skills Certified” discovers during the onboarding period that the candidate struggles with real-world application. Nobody is lying. Everyone invested in good faith. But the signal—the credential—isn’t predicting actual capability.

This gap between certification and demonstrable capability has become the hidden cost of Southeast Asia’s digital transformation. And for SMEs, it’s catastrophic.

The training paradox: Credentials without capability

Here’s what the data reveals: training completion is not the same as job readiness. The distinction matters more than we’ve admitted.

Researchers across multiple industries have documented this phenomenon. Cloud Range’s 2025 research on technical workforce readiness is unambiguous: “Knowledge is what you learn. Readiness is what you can perform—and those are not the same. In a live incident, the difference between knowing what to do and being able to execute in real time under uncertainty is dramatic.”

This isn’t a criticism of training programmes. It’s a description of a fundamental gap between learning and performance.

Consider Google’s experience, documented by Cornerstone OnDemand. For years, the company screened job candidates using traditional credentials: transcripts, GPAs, test scores. After hiring thousands of people, Google researchers concluded these credentials were essentially “worthless” for predicting actual job performance. Only 43 per cent of workers in STEM roles even possess STEM degrees—yet those roles are filled nonetheless, suggesting that credentials and actual capability are loosely correlated at best.

In Southeast Asia, this gap has been replicated at scale. The SHRM Global Worker Project (2025) found that globally, 37 per cent of workers hold jobs that don’t align with their skills, while 53 per cent report their roles don’t match their education and training. But the regional data is more alarming: Singapore’s Ministry of Manpower and National Trades Union Congress (NTUC) study (2025) found that hiring challenges are increasingly driven by “skills specificity rather than qualification mismatches”—meaning employers aren’t struggling to find people with credentials; they’re struggling to find people with demonstrated expertise in the specific capability needed.

Translation: The credential exists. The capability doesn’t.

Also Read: “The AI did it” is not a defence; it is a confession

The hidden cost: What credential-capability mismatch actually costs

When certification becomes divorced from capability, three cascading problems emerge for organisations, particularly for resource-constrained SMEs.

• First, hiring decisions fail silently. An SME manager reviews a resume showing “AI Fundamentals Certified.” The hiring process validates the credential. The candidate onboards. Within weeks, the manager realises the person can apply frameworks in training conditions but freezes when facing real systems. The hire was made on a false signal—and SMEs, lacking large HR infrastructure, often don’t have backup plans or retraining budgets.
• Second, organisational anxiety increases. When 24.3 per cent of Singapore employers report experiencing skills gaps in their workforce, and 49.9 per cent report this causes increased workload for other staff, you’re describing a system where “trained” people can’t actually perform, forcing colleagues to compensate. The trained employee feels inadequate despite their certificate. Their manager feels misled by the training system. The organisation’s confidence in development programmes erodes.
• Third, competitive advantage evaporates. SMEs are racing to adopt AI to compete with larger rivals. But if their hiring signal—the credential—doesn’t predict whether someone can actually build AI systems, deploy models, or integrate AI into operations, they’re hiring randomly and hoping. In a competitive market, hope is a business risk.

This is where the problem reveals itself as a systems issue, not an individual or training-quality issue.

The signal integrity problem: Why credentials fail in APAC

Southeast Asia’s workforce development system has optimised for measurable completion metrics rather than capability verification:

What gets measured:

• Training volume (555,000 trained annually in Singapore alone)
• Certification completion rates
• Course quality and practical components

What doesn’t get measured:

• Can the certified person actually perform on the job?
• Do credentials predict job success, retention and performance?
• Is the certification signal reliable?

The result is a market-wide problem. When 16 per cent of specialised professional, manager, executive, and technician (PMET) roles in Singapore remain unfilled for six or more months, employers specifically cite difficulty finding people with demonstrated technical expertise—not people with credentials.

The credential system hasn’t failed because the training is poor. It’s failed because certifications and actual capability are being treated as equivalent when they’re not.

Also Read: The most sophisticated AI strategy is a puzzle hunt in Toa Payoh

The AI-powered enterprise solution: Bridging signal integrity

This is where AI-powered enterprise solutions become the game-changer for SMEs in Southeast Asia.

Traditional hiring systems can filter for credentials. They struggle to verify capability. AI-powered assessment platforms can do what neither training programmes nor conventional recruitment can: assess demonstrated capability—not just knowledge of frameworks—at scale and with consistency.

These solutions work by distinguishing between three different assessment layers:

• First, deterministic signals: Keyword and semantic analysis identify formal qualifications and technical vocabulary. Someone who says they “trained in Python” appears here. But this doesn’t prove they can debug production code under pressure.
• Second, semantic understanding: Advanced models evaluate whether someone can explain concepts in their own words, suggesting deeper comprehension than memorisation. This is closer to capability but still incomplete.
• Third, capability assessment: This is the layer most SMEs lack access to. AI-powered capability assessment goes deeper: Can this person actually do the work? Can they apply knowledge to novel problems? Can they integrate with existing systems? Will they perform in real conditions?

For SMEs, this third layer is transformative. A small team can now make hiring decisions with the same rigour a large enterprise could afford through expensive assessment centres. An SME can distinguish between “certified” and “actually capable” before hiring. They can identify which trained employees are genuinely ready for deployment in AI initiatives.

The competitive imperative for SMEs

SMEs in Southeast Asia face a unique time constraint. Larger competitors are adopting AI faster. Regulatory environments (EU AI Act, Japan’s ¥10 trillion Trustworthy AI 2030 mandate) are tightening requirements. The window to build AI-ready capability is closing.

But SMEs can’t afford to hire and fail repeatedly. They don’t have the budget to train an entire team, discover half aren’t capable, and retrain. They need to know, before hiring or promoting, whether their team members actually have the capability that their credentials claim.

Also Read: How AI is dismantling the risk pool in insurance

AI-powered enterprise assessment solutions solve this by:

• Reducing mis-hire costs: Verify capability before hiring, not after onboarding failure
• Optimising training ROI: Identify which trained employees are genuinely ready for deployment
• Accelerating AI adoption: Deploy capability with confidence rather than guessing
• Building organisational trust: When capabilities are verified, teams move faster and with less anxiety

The game-changer moment

We’re at an inflection point. Southeast Asia has solved the training problem—the region demonstrates this daily with millions of course completions. What remains unsolved is the verification problem: reliably determining who actually has capability versus who has certification.

SMEs that address this first—that adopt AI-powered enterprise solutions to verify demonstrated capability rather than relying on credentials—will outcompete peers who continue hiring blindly. They’ll deploy trained talent more effectively. They’ll build confidence in their teams. They’ll accelerate their competitive position.

The credential-capability gap that seemed like a training problem is actually an assessment and verification problem. And for the first time, AI-powered enterprise solutions make that verification affordable and scalable for organisations of any size.

That’s the game-changer Southeast Asian SMEs have been waiting for.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

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

The post The real workforce challenge: Bridging the credential-capability gap appeared first on e27.

Artificial intelligence is no longer sitting at the edge of enterprise experimentation. Across the Asia Pacific, AI assistants and autonomous agents are moving into live business environments, embedded across email, customer support, internal messaging, cloud applications and collaboration workflows.

That shift is creating an enormous opportunity. AI can help organisations move faster, automate routine work, improve customer experience and support better decision-making. But it is also changing the security equation. As AI becomes part of how work gets done, it is expanding where risk appears, how quickly incidents move, and how difficult it is for security teams to investigate what happened.

Proofpoint’s 2026 AI and Human Risk Landscape report shows that AI adoption in Singapore has already moved well beyond the pilot stage. 87 per cent of organisations in Singapore have deployed AI assistants beyond the pilot stage, and 70 per cent are actively piloting or rolling out autonomous agents. Yet security readiness has not kept pace. Close to three-fifths of these organisations describe their AI security posture as catching up, inconsistent or reactive. 38 per cent have already experienced a suspicious or confirmed AI-related incident.

This is the gap that should concern security leaders. AI is not waiting for governance frameworks to mature. Security leaders in Asia are under more pressure to address key areas of concern.

AI has expanded the attack surface

For many years, cybersecurity strategies were built around familiar control points: email, endpoints, cloud applications, identities and data repositories. Those still matter. But AI is now connecting these environments in new ways, allowing risk to move across workflows at machine speed.

In Singapore, email remains the most common AI-related threat vector, affecting 58 per cent of organisations. But exposure now extends much further: SaaS and cloud applications at 44 per cent, AI assistants or agents at 41 per cent, and collaboration tools such as Teams or Slack at 44 per cent. Among organisations that experienced an AI-related incident, exposure rises across every channel, including 61 per cent in file sharing platforms and 58 per cent involving collaboration tools.

Also Read: Singapore’s data analysts trust AI to work, not to think

This matters because enterprise work no longer happens in a single channel. A sensitive document may move from email into a collaboration platform, be summarised by an AI assistant, stored in a cloud application, and referenced by an autonomous workflow. Each step creates another point where data, identity and intent need to be understood.

Many organisations already have some forms of AI security controls, for example, monitoring shadow AI applications. However, the critical visibility is whether those controls can see across the connected environment how AI is actually being used.

Data security and AI security are the same problem

One of the most common structural errors in how organisations approach AI security is treating it as a separate workstream from data security. It is not. They are facets of the same problem, and solving one without addressing the other creates compounding exposure.

The earliest AI security challenge was clear: employees were using consumer AI tools to process sensitive business information. In 2025, 63 per cent of employees who used AI applications uploaded confidential company data, such as source code and customer records, to personal chatbot accounts. According to IBM’s Cost of a Data Breach Report, shadow AI breaches cost an average of US$670,000 more than standard security incidents, driven by delayed detection and difficulty determining the scope of exposure.

The second wave is more complex. As organisations moved to enterprise AI platforms — Microsoft Copilot, Salesforce Einstein, and others — the question became not whether data was leaving the organisation, but whether AI tools were accessing only the data they were supposed to. That is a data security problem expressed through an AI lens.

The third wave is real-time and agentic. Autonomous agents do not just respond to prompts. Similar to humans, they connect to external tools and MCP servers, acquire new capabilities, and act on data across connected systems. Understanding what an AI agent is doing requires capturing not just the prompt and response, but every tool call and downstream action in between. When security teams do not have visibility into what AI is connecting to and acquiring, they cannot tell the board they have it under control.

Also Read: AI-powered business automation: How SMEs are transforming operations in Southeast Asia

Gartner projects that by the end of 2026, up to 40 per cent of enterprise applications will integrate with AI agents, up from less than five per cent in 2025. It also predicts that by 2028, 25 per cent of all enterprise GenAI applications will experience at least five minor security incidents per year, up from nine per cent in 2025. The risk is scaling faster than governance.

Security and data governance teams need a shared view: what data exists, who and what has access to it, and how AI agents are actually using it. Having a clear view of all your data is not fictional, and it should be the foundation of building robust AI security for any organisation.

Tool sprawl is holding security teams back

Fragmented security stacks are compounding the challenge. Almost all organisations in Singapore say managing multiple security tools is at least moderately challenging, and 61 per cent describe it as very or extremely difficult. Respondents cite operational cost pressures, integration challenges and difficulty correlating threats.

When controls sit in separate systems, security teams lose time moving between dashboards, reconciling alerts and trying to connect activity across email, cloud, collaboration and AI systems. That delay matters when incidents can spread across workflows quickly.

As AI scales, security architecture becomes a strategic priority. More than half of Singapore organisations are actively pursuing vendor and tool consolidation, and 58 per cent believe a unified platform is more effective than point solutions. This reflects a broader shift. Organisations are recognising that AI security cannot be solved with isolated controls. It requires an architecture that can protect people, data and AI systems across the channels.

AI adoption in Singapore and Asia Pacific is not slowing down. The boards and CEOs driving it are right that falling behind carries a real competitive cost. The security leaders are now in a perfect position to enable this AI innovation with the visibility to secure it, govern it, and defend it. That is what setting the pace looks like.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

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

The post AI is already inside the enterprise. Has security kept up in Asia? appeared first on e27.

Vietnam’s electric vehicle (EV) market is entering a more practical phase. After years of attention on vehicle launches, subsidies, and consumer adoption, the next question is becoming harder to ignore: where will all these cars and motorbikes charge, and who will pay to build the network?

That question sits at the centre of Grab’s latest move in Vietnam. On July 10, the Singaporean superapp company announced a direct investment in EBOOST, a Vietnamese electric vehicle charging platform and infrastructure provider.

Also Read: Why rising fuel costs are pushing drivers towards EVs

The investment size was not disclosed, but the deal is notable for what it says about the next stage of EV adoption in Vietnam: charging is no longer just a real estate or infrastructure problem. It is becoming a platform business tied to mobility, payments, driver economics, and fleet utilisation.

The investment follows a memorandum of understanding signed by the two companies last November. At that point, the partnership was framed around giving Grab’s driver-partners easier access to EBOOST’s charging network. The new investment deepens that relationship and gives EBOOST additional financial capacity to expand its footprint across Vietnam.

Before the Grab partnership, EBOOST had already built one of the country’s larger independent charging networks, with more than 2,500 charging points and over 10,000 EV users nationwide. Its charging locations cover office buildings, residential developments, public destinations, parking facilities, and other everyday sites where vehicle downtime can be turned into charging time.

For Vietnam, this kind of distributed network matters. Unlike markets where charging infrastructure is concentrated along highways or in suburban homes, Southeast Asian cities are denser, more fragmented, and heavily reliant on two-wheelers. A successful EV charging strategy must serve office workers, apartment residents, ride-hailing drivers, delivery riders, and taxi fleets, often in the same neighbourhood but with very different charging habits.

From MoU to in-app charging

The clearest sign of the partnership’s commercial value is the integration of EBOOST’s network into the Grab Driver app.

Since April, Grab-Car driver-partners have been able to use the app’s EV Charging feature to find nearby EBOOST stations, start charging sessions, and complete payments without switching platforms. Hundreds of charging points have already been connected to the system.

Also Read: Electrifying Southeast Asia: Unleashing the radical potential of electric vehicles

That may sound like a product detail, but it addresses a real barrier for drivers. Charging is not just about the price per kilowatt-hour. It is also about route planning, waiting time, payment friction, reliability, and confidence that a charger will be available when needed. For ride-hailing drivers, every extra minute spent hunting for a charger is potential income lost.

Early usage data suggests the service is finding a repeat audience. More than 70 per cent of driver-partners continue using the service within the first seven days. For EBOOST, that points to stronger charger utilisation and recurring revenue. For Grab, it helps make EV use more practical for drivers whose daily income depends on predictable vehicle uptime.

The next phase will extend the same charging experience to electric motorbike driver-partners. That could be more consequential than the car segment alone. Vietnam remains one of the world’s largest motorbike markets, and the electrification of two-wheelers will be central to any meaningful shift in urban transport emissions.

Integrating thousands of additional charging points for motorbike users could give Grab a stronger role in shaping driver behaviour at scale.

Why Grab needs charging partners

Grab’s interest in EV infrastructure is not surprising. Across Southeast Asia, ride-hailing and delivery platforms face growing pressure to reduce emissions, while drivers remain highly sensitive to operating costs. EVs can lower fuel and maintenance expenses, but only if charging is convenient, affordable, and reliable.

That is where charging operators such as EBOOST become strategically important. A platform can encourage drivers to switch to EVs, but it cannot afford a poor charging experience that disrupts earnings. By embedding charging access into its driver app, Grab can reduce friction for drivers while gathering data on demand patterns, station performance, and charging behaviour.

Also Read: Grab’s US$600M deal could save Taiwan from a delivery monopoly

The model also reflects a broader shift in EV infrastructure. In early markets, charging networks were often built as standalone assets. In more mature ecosystems, they are increasingly tied to software layers: booking, payments, fleet management, energy optimisation, loyalty, and data analytics. EBOOST’s proprietary software platform is designed to support both electric cars and motorbikes, giving it room to serve mixed fleets and different user groups.

For Vietnam, that flexibility is important. The country’s EV market is not moving in a straight line. Private car adoption, taxi electrification, delivery fleets, e-motorbikes, and public charging demand are developing at different speeds. Charging companies that can serve multiple vehicle types may be better placed than those built around a single use case.

A crowded but still-open market

EBOOST is not building in an empty field. Vietnam’s EV charging landscape is shaped heavily by VinFast and its related infrastructure ecosystem, particularly V-Green, which has been expanding charging access to support the country’s largest domestic EV manufacturer. Regional players are also watching the market closely, including Singapore-based Charge+, which has been building cross-border charging ambitions in Southeast Asia, and other energy and mobility companies exploring EV infrastructure across the region.

The competitive question is whether independent networks can create enough utilisation outside manufacturer-led systems. EBOOST’s partnership with Grab gives it one potential answer: aggregate demand through a large mobility platform rather than relying only on walk-in consumer charging. If Grab’s EV driver base grows, EBOOST could benefit from more predictable charging volumes, while Grab gains a charging layer without having to build and operate the entire network itself.

That matters because EV charging is a capital-intensive business. Hardware deployment can be expensive, site acquisition is complex, and payback periods depend heavily on utilisation. A charging station in the wrong location can sit underused; one tied to a reliable flow of commercial drivers can become a recurring revenue asset.

The Southeast Asian test case

Vietnam is emerging as one of Southeast Asia’s most closely watched EV markets. Its combination of urban density, motorbike dependence, local manufacturing ambition, and fast-growing digital services makes it a useful test case for the region. If companies can solve charging access for ride-hailing cars and motorbikes in Vietnam, similar models could be adapted in Indonesia, Thailand, and the Philippines.

Also Read: Grab’s US$425M Stash acquisition is about AI coaching, not America

Still, execution will decide the outcome. EBOOST will need to expand without sacrificing reliability, while Grab must ensure the economics work for drivers, not only for platform targets. Charging access has to be priced and located in ways that make daily use practical.

The investment gives EBOOST stronger backing at a time when Vietnam’s EV ecosystem is moving from headline ambition to operational detail. The next contest will not be won only by who installs the most chargers. It will be won by the companies that make charging invisible enough for drivers to build their working day around it.

The post Grab invests in EBOOST as Vietnam’s EV charging race shifts into higher gear appeared first on e27.

Bitcoin sits at US$66,188.15 this morning, down 0.65 per cent over the past 24 hours, and the number tells only a fraction of the story. The cryptocurrency pulled back from a one-month high near US$67,000 as WTI crude oil surged above US$85 and then climbed further toward US$94 per barrel amid an escalating US-Iran conflict now in its second week.

The broader crypto market cap slipped 0.47 per cent alongside Bitcoin, volume contracted 6.14 per cent, and Bitcoin dominance held near 59 per cent with no meaningful capital rotation into altcoins. On the surface, this looks like a gentle consolidation after a recovery rally from July lows.

But when I step back and look at the full macro picture unfolding on July 23, 2026, I struggle to share the optimism that we have found a floor. The global factors stacking up right now suggest Bitcoin has more downside to endure before any sustainable recovery takes shape.

The oil story dominates everything else at this moment. Crude prices above US$94 per barrel represent the highest levels since June, and they carry direct implications for inflation expectations and Federal Reserve policy. The US-Iran conflict shows no signs of de-escalation as it enters its second week, which means supply disruption risks remain firmly on the table. Higher energy costs feed into transport, manufacturing, and consumer prices across the board. Airlines already face margin compression.

Treasury yields hover near 2026 peaks as bond markets price in the possibility that the Fed keeps rates higher for longer. The July 28 FOMC meeting looms as the next critical catalyst, and if the statement hints at delayed rate cuts or, worse, another hike, risk assets, including Bitcoin, will absorb the blow directly. Bitcoin in this environment behaves exactly like a leveraged tech stock rather than a decoupled store of value, and that correlation works against holders when macro conditions deteriorate.

Also Read: Bitcoin just broke US$66,000: Is this the start of the next bull run or a trap for late investors?

The equity backdrop reinforces my caution. Wall Street benchmarks finished lower overnight, with S&P 500 and Nasdaq futures slipping as megacap tech earnings delivered mixed signals. Tesla fell around four per cent after missing both revenue and margin estimates. Alphabet reported strong Q2 cloud growth but slid in extended trading as investors baulked at plans to increase capital expenditures. Yes, Super Micro Computer surged 20 per cent on strong AI server margin forecasts, but that single bright spot does not offset the broader disappointment.

The AI narrative that has propped up markets for over one year now faces scrutiny on whether spending translates into returns. When growth stocks wobble, Bitcoin wobbles harder. The 6.14 per cent drop in crypto trading volume suggests buyers have stepped back and lack the conviction to defend current levels. This is not the behaviour of a market that has found its bottom.

Technically, Bitcoin tests its daily pivot near US$66,103 right now. The immediate Fibonacci support sits at US$64,750, representing the 23.6 per cent retracement level. Below that, the US$63,000 to US$63,400 zone serves as the next meaningful floor. Overhead, the 100-day EMA near US$68,000 caps any rally attempt. The structure looks neutral on paper, but I read it as fragile.

A close below US$64,750 opens the trapdoor toward US$63,000, and given the macro headwinds I just described, I think that break becomes more probable with each passing day of elevated oil prices and unresolved geopolitical tension. The recovery channel from July lows remains intact for now, but channels break, and they tend to break in the direction of the prevailing macro wind.

Also Read: Bitcoin reclaims key technical levels, Ethereum leads broader market gains

Grayscale research head Zach Pandl offers a more constructive view, suggesting Bitcoin’s recent price low might hold if the Federal Reserve ends interest rate hikes and economic growth remains stable. He treats Bitcoin as a mature asset influenced by growth and Fed policy, rather than by the traditional four-year cycle model, which would predict a longer bear market and deeper declines.

Grayscale also points to the CLARITY Act and Strategy’s improved financial position, including a US$216 million Bitcoin sale that strengthened cash reserves and reduced forced selling risks, as structural positives. I respect that framework, but it relies on the Fed cooperating and growth holding steady. With oil above US$94 and inflation concerns resurfacing, the Fed has every reason to stay hawkish. The conditions Pandl requires for a bottom simply do not exist right now.

SkyBridge founder Anthony Scaramucci argues that Bitcoin will grind higher from here and cannot get much worse. I appreciate the sentiment, but the global picture tells a different story. Mixed Asian markets preparing for a cautious open, a steady US Dollar against the Yen and Euro that signals continued risk aversion, and geopolitical tensions with no resolution timeline all point toward sustained pressure. The AI boom cushions some of the blow in equities, but it does not immunise crypto from a liquidity squeeze if yields push higher.

Here is where I land. Bitcoin at US$66,188.15 reflects a market in pause, not a market in recovery. The 0.65 per cent daily decline understates the vulnerability beneath the surface. Oil above US$94, an active US-Iran conflict, disappointing tech earnings, yields at 2026 peaks, and an FOMC meeting five days away create a cocktail of risk that has not fully priced into crypto.

The 6.14 per cent volume decline confirms that participants are waiting on the sidelines rather than accumulating. Bitcoin dominance at 59 per cent shows no rotation, no excitement, no fresh capital entering the ecosystem. I do not think we have bottomed.

The US$64,750 level will face a serious test before July ends, and if the FOMC disappoints or oil pushes toward US$100, the US$63,000 zone becomes the realistic near-term target. Patience, not optimism, serves holders best in this environment. The macro picture has not given us permission to call a bottom, and until it does, every rally toward US$68,000 looks like a selling opportunity rather than a breakout.

—

The post Bitcoin volume drops 6.14%, and everyone calls a bottom, I disagree appeared first on e27.

Deep tech in a fractured world needs something very different.

For much of the last 30 years, the working assumption behind technology and capital was simple: the world was converging into one increasingly integrated market. If you could build a product that scaled, global demand and global capital would be there to meet you.

That assumption is now clearly breaking down. Supply chains are being rewired, export controls are spreading, and critical technologies are being treated as instruments of statecraft rather than just engines of growth. The question for Southeast Asia is whether it wants to be a spectator to this shift, or a protagonist.

Deep tech sits right in the middle of this story. It is capital intensive, politically sensitive, and deeply entangled with physical infrastructure and long-term industrial policy. Yet many of the funding models we rely on were designed for asset‑light software, not for advanced manufacturing, new energy systems, or frontier materials.

If Southeast Asia wants a meaningful role in this new order, it cannot rely solely on importing technology and exporting talent. It has to build its own deep tech platforms – and it has to do so with an investment model that acknowledges fragmentation rather than assuming frictionless global markets.

The deep tech paradox

There is a paradox at the heart of deep tech today.

On one hand, governments and corporates worldwide describe it as strategically important. Climate solutions, AI for science, semiconductors, and advanced manufacturing all sit near the top of policy agendas. In Southeast Asia, reports like DealStreetAsia’s The State of Deep Tech in SE Asia 2025  note that deep tech’s share of overall funding is rising, even as absolute capital fell during the recent funding winter.

On the other hand, a lot of the IP that could underpin these sectors still struggles to leave the lab. High‑value patents and prototypes often stall in what investors like to call the “valley of death”: that messy, expensive space between proof of concept and commercial scale.

Traditional venture capital evolved around “optionality”: spread small cheques across many companies, keep ownership light, and hope a handful of outliers carry the fund. That logic made sense when the product was software you could ship globally at marginal cost. It is misaligned with deep tech, where outcomes depend on engineering discipline, regulatory engagement, and long-term offtake contracts.

Also Read: Deeptech’s secret: Ignore the market, master the engineering, and let opportunity find you

In a fracturing world, that misalignment becomes more dangerous. Export controls, national security reviews, and shifting sustainability rules can redraw a company’s viable markets overnight. Treating deep tech as a spray‑and‑pray portfolio of lottery tickets is no longer just inefficient; it increases the risk that strategically important IP never reaches scale at all.

From exposure to control

Fragmentation doesn’t just increase risk. It also changes what “good” looks like for investors and builders.

In a flat world, the main question was often, “How do I maximise exposure to a theme?” In a fractured one, the more relevant question becomes, “Where do I need real control – over governance, capital structure, supply chains, and commercialisation?”

In Southeast Asia, a new pattern is emerging in that “messy middle” between traditional venture capital and private equity. Instead of spreading capital thinly, some platforms are taking significant stakes in a small number of ventures, combining capital with operating control, and standardising parts of the commercialisation process.

A key design choice is to start at higher Technology Readiness Levels – TRL 7 to 9 – where core scientific risk has already been resolved through public–private research ecosystems. In Singapore, for example, institutes such as ASTAR and university labs have built a deep pipeline of such IP, and recent work by McKinsey, the Singapore Economic Development Board (EDB) and Tech in Asia in AI in Southeast Asia: An era of opportunity shows how AI and related technologies are moving beyond pilots into scaled deployment.

By entering at this stage, investors and operators can focus on market design, go‑to‑market architecture, and capital efficiency rather than basic feasibility. Just as importantly, they can design governance and cap tables from the outset, which matters when regulatory and geopolitical risks are as material as technological ones.

Deep tech as a “non-aligned” asset class

In this environment, it’s helpful to think of deep tech as a potential “non‑aligned” asset class.

The most valuable technologies of the next decade – from advanced manufacturing and energy systems to critical materials – are likely to be contested by multiple blocs, rather than dominated by a single geography. Companies structurally tethered to one jurisdiction or standard can find their freedom to operate constrained as policies shift.

By contrast, platforms that anchor IP and governance in trusted hubs, while diversifying markets and manufacturing across regions, can become shared infrastructure rather than instruments of any one industrial strategy.

Southeast Asia, and Singapore in particular, is unusually well positioned to build such platforms. The region sits at the intersection of US, Chinese, and regional supply chains. Singapore offers a credible legal and regulatory environment, and its AI and tech ecosystems are maturing quickly. The AI in Southeast Asia: An era of opportunity report, for example, finds that nearly half of companies surveyed in the region have moved beyond AI pilots, putting Southeast Asia ahead of the global average. A Business Times summary notes that more than 80 per cent of companies are already piloting and scaling AI projects.

Anchoring IP in Singapore while designing ventures that can route production and customers across Asia, Europe, and beyond is one way to turn fragmentation into optionality. In practice, that means thinking early about export controls, dual‑use risks, data localisation, and AI governance frameworks such as ASEAN’s AI governance guide and Singapore’s Model AI governance guidelines.

Also Read: Why traditional marketing fails for complex B2B and deeptech products

From discovering to industrialising

The deeper shift, though, is recognising where the real bottleneck lies.

We are no longer constrained primarily by a lack of scientific discovery. Labs around the world – including those in Southeast Asia – are full of promising high‑TRL IP. The real constraint is institutional: our ability to take that IP and industrialise it, turning it into companies with credible revenue, governance, and liquidity paths.

Some emerging platforms treat company building explicitly as an engineering problem. They standardise finance and governance templates, regulatory pathways, and operational playbooks, and apply these across a concentrated portfolio where they hold meaningful ownership from inception.

Rather than backing dozens of experiments, they co‑build a smaller number of high‑conviction ventures, often with the aim of reaching public markets within a defined timeframe. Exchanges such as SGX, HKEX, and NASDAQ are already home to advanced manufacturing and deep tech listings. EDB’s Destination Southeast Asia 2024 report shows how the region’s tech hubs are attracting more sophisticated capital, while DealStreetAsia’s deep tech reviews highlight a growing share of deep tech deals in the overall venture mix, even after a pullback in funding.

In a fracturing world, this approach has two advantages. It keeps cap tables and governance relatively clean, which simplifies regulatory engagement and cross‑border partnerships. And it gives investors a clearer line of sight to liquidity, which matters when global IPO windows are more volatile, and capital is becoming more selective.

Where Southeast Asia fits

All of this brings us back to the original question: where does Southeast Asia fit in a fracturing world?

On one level, the region is another theatre in a global competition for capital, talent, and supply chains. On another, more interesting level, it can be a builder of the deep tech platforms that fragmentation actually requires: resilient, multi‑market, and anchored in trusted institutions.

If Southeast Asia can consistently take high‑quality IP from its research institutions and partners, industrialise it, and bring it to market with credible governance and liquidity, it becomes more than a manufacturing base or testbed. It becomes a generator of infrastructure‑grade deep tech – platforms that multiple blocs can depend on, but none can easily dominate.

That is ultimately how the region’s voice becomes part of the conversation, rather than reacting after the fact: not by recreating Silicon Valley’s venture playbook, but by building the kinds of deep tech institutions that a fracturing world will increasingly need.

—

The post Deeptech and a fracturing world: Why Southeast Asia needs a new playbook appeared first on e27.

Most leaders entering a new market think first about demand. They ask whether the problem is large enough, whether the timing is right, whether regulation is favourable, whether distribution can be acquired at acceptable cost, and whether the economics can support scale. Those are sensible questions, but they often arrive too early. Before a market can scale, before it can standardise, before it can attract sustained capital and serious institutional participation, it has to solve something more basic. It has to establish a trust primitive.

By trust primitive, I do not mean brand warmth or a vague sense of confidence. I mean the foundational mechanism that allows strangers, institutions, and counterparties to participate despite uncertainty. It is the smallest reliable unit of belief that makes the market usable. In some categories, that primitive is escrow. In others, it is identity verification, a guarantee, a clearing mechanism, transparent pricing, dispute resolution, regulatory oversight, or auditability. The exact form changes by sector, but the strategic truth does not. Every new market becomes real only when participants know what they can rely on, what happens when something goes wrong, and who absorbs the consequences.

That is why so many markets look promising in theory and fragile in practice. 

New markets do not fail because of weak demand

A surprising number of early market failures are misdiagnosed. We often say customers were not ready, adoption was too slow, or the proposition was not compelling enough. Sometimes that is true. But in many cases, the real problem is that the market asks people to take too much on faith.

When a market is new, uncertainty exists at every layer. Buyers do not know whether quality claims are real. Sellers do not know whether they will be paid fairly or on time. Partners do not know whether standards will hold. Regulators do not know whether risks are visible early enough. Investors do not know whether apparent growth is durable or merely subsidised experimentation. In that environment, even a strong product can struggle because the surrounding conditions are too ambiguous for meaningful commitment.

This is where strategy often gets superficial. Teams focus on proposition design, pricing, or acquisition before addressing the deeper question of assurance. What would make a rational participant comfortable enough to depend on this market, not just sample it? That question sounds softer than it is. In reality, it is structural. 

Trust is not a brand outcome; it is market infrastructure

Consumers may say they trust a brand, but what they often mean is something more concrete. They believe payments will settle correctly. They believe data will be handled properly. They believe there is recourse if something goes wrong. They believe quality has been checked by someone other than the seller. They believe abuse will be contained. They believe the rules will be applied consistently. In other words, what looks like trust is often confidence in invisible infrastructure.

Also Read: How creativity, commerce and AI collide in mid-2026 marketing mix

This matters because markets do not stabilise through aspiration alone. They stabilise through mechanisms that reduce the cost of belief. That may include insurance, certification, guarantees, identity systems, standard contracts, transparent governance, independent oversight, or rules around loss allocation. Once these mechanisms are in place, the market no longer depends on every participant making a heroic judgment call every time they engage. The system does more of the work.

The first trust primitive is rarely the final one

Another mistake is assuming trust is solved once a market gets early traction. In reality, trust evolves in stages, and the primitive that unlocks early adoption is often different from the one required for institutional maturity.

Early consumer platforms, for example, often rely on visible signals such as reviews, ratings, social proof, and simple guarantees. Those mechanisms can be enough to establish initial confidence among retail users. But once the market seeks enterprise adoption, regulatory approval, or critical mass across a more complex value chain, those same mechanisms become insufficient. Institutions do not make decisions on the basis of community sentiment. They want process controls, audit trails, contractual clarity, governance standards, measurable accountability, and credible remediation.

Every serious market solves the question of loss

If I had to reduce the trust primitive to a single test, it would be this. When something fails, who carries the loss, and how quickly is that answer known?

This is where abstract conversations about trust become concrete. Markets that scale are not markets without failure. They are markets where failure is legible, containable, and allocable. Participants know the boundary conditions. They know whether a transaction can be reversed, whether liability sits with the platform or provider, whether disputes can be adjudicated, whether fraud is insured, whether records are accepted as evidence, and whether harm can be corrected without destroying participation.

This is one reason payments matured through rules, networks, chargeback mechanisms, and settlement disciplines. It is why financial services depend so heavily on supervision, capital requirements, complaints handling, and conduct frameworks. It is why digital identity remains such a hard problem in many emerging categories. It is why AI markets will increasingly be judged not only by capability, but by traceability, explainability, and responsibility when decisions cause harm.

The best growth strategy is often trust architecture

A well-designed trust primitive compresses adoption friction. It shortens decision cycles. It reduces the burden on frontline sales teams to overexplain risk. It lowers compliance anxiety. It improves repeat behaviour because participants are not renegotiating uncertainty every time they return. Most importantly, it changes the shape of the market itself. More counterparties become willing to join, more workflows can move from exception handling into standard process, and more capital becomes comfortable backing long-term participation.

Also Read: The future of marketing isn’t about AI, it’s about judgment

This is why the most consequential strategic moves in a new market often look unglamorous from the outside. They involve rule setting, standard creation, liability design, governance forums, audit models, certification systems, customer protections, and interoperable controls. These are not usually celebrated as growth stories in the early narrative. But they are precisely what separates a category that remains interesting from one that becomes durable.

The paradox is that trust architecture can feel like friction in the short term while creating expansion in the long term. Weak leaders avoid it because it slows the initial story. Strong leaders invest in it because it changes the ending.

The first question should not be market size

When evaluating a new market, the smartest first question is not how large it could become. It is what participants need in order to trust it enough to rely on it.

That framing changes the quality of strategic thinking. It moves the conversation away from enthusiasm and towards structure. It forces clarity on institutions, incentives, safeguards, and failure management. It also reveals whether the company is actually building a market or merely exploiting a temporary gap before trust catches up with reality.

This matters especially for leaders trying to build new businesses in complex sectors. The closer a market is to money, identity, data, safety, or operational continuity, the less room there is for trust to remain informal. In these domains, trust must be engineered, evidenced, and governed. Without that, scale tends to arrive before legitimacy, and that is usually when the real problems begin.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

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

The post The evolution of trust: From early adoption to institutional maturity appeared first on e27.