I keep a small notebook by my bed. Not for ideas. Just things I don’t want to forget.

Last week I wrote: “What does this place look like in 2040?”

Not for a deck. Not for a client. I don’t have kids. The usual anchors people use to think about the future, school fees, inheritance, legacy through bloodline, none of that applies to me. So I find other anchors. Climate is one of them. It’s personal in a way that’s hard to explain at a conference but easy to feel at midnight.

I’ve been watching SEA climate capital closely for a few years now. And a number in a recent Tracxn report stopped me cold.

Over 900 backers have put money into climate tech in SEA. You can count the ones who kept showing up on one hand.

Let that sit for a second.

The crowd that showed up once

900 is an impressive number until you look at what it actually means. One check. One conference announcement. One ESG box ticked somewhere in an LP deck. Then back to whatever they were doing before.

The ones who kept showing up are SEEDS Capital, Entrepreneur First, 100×100 (formerly Wavemaker Impact, rebranded June 2026 after spinning out as an independent fund manager with a fresh US$100M mandate to build 50 climate companies across SEA and India), SGInnovate and East Ventures. Between them, they’ve done the unglamorous work of showing up round after round across companies that haven’t yet become household names.

That’s not a coincidence. That’s conviction. And conviction in climate is rare because the timeline is long, the exits are slow, and the narrative keeps shifting between optimism and panic depending on what’s burning that week.

The others weren’t lying about caring. They just weren’t prepared for what caring actually costs in this space.

Also Read: Investing in impact: High-growth tech for climate and community

What the data says about where the money went

The top five funded sectors are all infrastructure plays. Solid waste, smart grid, energy efficiency, air pollution, renewable energy. The top five funded companies are almost all in electric mobility. Beam, Neuron, Dat Bike, ALVA. Hard assets, visible units, clear revenue models.

This isn’t surprising. Capital backs what it can underwrite. Infrastructure and mobility fit the frame. What doesn’t fit — adaptation, resilience, the harder climate problems without a clean SaaS analogy — stays unfunded.

Geographically, three-quarters of all climate capital went to Singapore. Indonesia and Vietnam are emerging, but the map is thin outside the main hubs.

2026 so far has been even quieter. Four rounds. US$17M. Down 59 per cent from the same period last year.

The crowd didn’t just come once. Some of them have already left.

What the data can’t see

Here’s the thing about a report that tracks 900 backers. It only sees what gets disclosed. Formal rounds. Announced deals. Tracxn does this well. But there’s a whole layer of capital doing real ground-level work that never shows up in any database.

Funds like Bali Investment Club, Indonesia-focused, impact-first, backing waste, agritech and climate ventures at the earliest stages in a market most Singapore-based funds fly over, won’t show up in the top five. They’re not doing 10 rounds in the data. They’re doing the first round in places nobody else is looking.

That gap between what the data captures and what’s actually happening on the ground is where a lot of the real conviction lives.

Also Read: Climate tech’s shift from doing good to doing well

What the five who stayed understand

100×100 doesn’t just write checks. They co-found companies from scratch, sit inside them, find the first customers, and build the team. That’s a different level of commitment than a seed round from a platform investor diversifying into climate for a season.

The difference between the handful who stayed and the hundreds who didn’t isn’t access to data or deal flow. It’s tolerance for a long, uncomfortable, uncertain road. Climate doesn’t reward impatience. The founders building in this space know that. The capital that stays knows that too.

The capital that came once was looking for a climate story. The capital that stayed is building one.

The notebook question

I still haven’t answered what I wrote last week. What does this place look like in 2040?

It depends on which kind of capital wins. The kind that showed up for the photo. Or the kind that’s still here when there’s nothing to announce.

I know which one I’m watching.

—

The post Funded: I keep a notebook by my bed with one question about SEA climate appeared first on e27.

The current correction in the digital asset market reflects a structural shift in investor behaviour rather than a random price fluctuation.

Bitcoin recently fell by 0.84 per cent over a 24-hour period to settle at US$59,526.31, which slightly outpaced the broader market decline of 0.94 per cent. This synchronised downward movement highlights how tightly integrated crypto assets have become with traditional financial markets, demonstrating an 85 per cent correlation with the S&P 500 index.

Institutional capital is actively rotating out of digital assets and back into traditional equities, creating a profound liquidity drain. Last week, exchange-traded funds tracking spot Bitcoin experienced US$1.79 billion in net outflows, marking the second-largest weekly redemption phase since these financial products launched.

A single-day redemption of US$445 million occurred on June 26, which provided clear evidence that institutional investors are reducing exposure. Over a longer horizon, these funds shed roughly US$6 billion over six weeks, while adjacent market reports indicate that total exits reached approximately US$6.4 billion over a 30-day period. Consequently, total assets under management for these investment vehicles plummeted from US$105.32 billion down to US$81.83 billion within one month, demonstrating that the structural buying pressure that catalysed previous market rallies has completely reversed.

This aggressive capital flight directly coincides with a broader macroeconomic tightening cycle and mounting geopolitical risks. The Federal Reserve continues to maintain a hawkish stance, with officials projecting a median interest rate forecast of 3.8 per cent for 2026. These higher-for-longer interest rate expectations have consistently strengthened the dollar, which naturally dampens demand for speculative, risk-sensitive assets.

This macro pressure intensified as fragile ceasefire negotiations between the United States and Iran stoked fears of conflict escalation, prompting global market participants to seek safety in cash. Although equity futures staged a minor recovery on June 29 after both nations temporarily pulled back from military strikes, the prolonged period of regional tension has left energy markets on edge and dragged European indices down by an average of one per cent.

The combination of institutional selling and macroeconomic headwinds triggered an immediate unwinding of high-risk leverage within crypto derivatives markets. Over a recent 24-hour window, the market suffered US$44.96 million in total liquidations, with long positions accounting for an overwhelming US$39.77 million of that total. This rapid liquidation sequence forced the asset price below its critical 200-week moving average of approximately US$62,383, which technicians widely respect as a key long-term trend indicator. The steep decline means Bitcoin now trades roughly 30 per cent lower in 2026, leaving it roughly 50 per cent below its historical peak established in October 2025.

Also Read: Why the 4.1% PCE inflation print just turned crypto into a high beta risk asset

While the overarching market structure remains transitionally bearish, certain technical indicators suggest that the current selling pressure might be reaching a temporary exhaustion point. The 14-day relative strength index plunged to 30.7, placing the asset on the verge of deeply oversold territory.

This technical condition indicates that if the current support zone between US$58,000 and US$59,000 holds firm over the coming days, a short-term relief bounce toward the US$62,000 level could easily manifest. Conversely, a definitive break below the US$58,000 threshold would likely trigger a fresh wave of liquidations, risking a rapid cascade down toward US$56,000.

A sustainable market recovery depends entirely on a stabilisation of fund flows and an easing of macroeconomic pressures. The broader financial landscape is experiencing a massive rotation, with Wall Street shifting capital out of underperforming assets and certain mega-cap technology equities to fund small-cap firms and blue-chip sectors.

Within the technology sector itself, a distinct wedge has formed between software hyperscalers struggling with infrastructure costs and memory component manufacturers like Micron Technology, which recently surged to outpace Meta and Tesla in valuation. If the massive capital rotation into chip makers and artificial intelligence infrastructure slows down, or if the Federal Reserve delivers a more dovish policy signal, capital may eventually flow back into the digital asset space.

—

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 dangerous liquidation cascade waiting below the US$58,000 support threshold appeared first on e27.

Singapore-based medtech company Biobot Surgical has raised SGD 20 million (about US$15.6 million) in a financing round led by ClavystBio, with participation from ZIG Ventures, as it looks to expand global adoption of its robotic-assisted prostate cancer care platform and enter the US market.

The company develops Mona Lisa, a platform that combines robotic needle guidance with MRI-ultrasound image fusion for targeted prostate biopsy and focal ablation procedures. The system is designed for transperineal prostate interventions, an approach increasingly preferred by clinicians seeking to reduce infection risks and improve access to hard-to-reach lesions.

Also Read: The role of biotech in taking India from developing to developed

The raise puts Biobot among a small but growing group of Southeast Asian medtech firms attempting to move beyond domestic or regional deployment and compete in heavily regulated global healthcare markets. For Singapore, where state-linked capital and research institutions have long sought to turn biomedical research into commercial companies, Biobot’s next phase will be watched as a test of whether homegrown medical robotics can scale internationally.

A Singapore medtech moving into a larger global fight

Prostate cancer is among the most commonly diagnosed cancers in men globally, with more than 1.4 million new cases each year. In the US, around one in eight men will develop prostate cancer in their lifetime, according to the American Cancer Society.

Diagnosis has traditionally relied on transrectal ultrasound-guided biopsy, but the field has been shifting towards transperineal procedures, where the prostate is accessed through the skin between the scrotum and anus rather than through the rectum. The change is being driven by lower infection risk, better targeting of certain areas of the prostate, and improved compatibility with image-guided and robotic workflows.

This is the clinical shift Biobot is betting on. Its Mona Lisa platform supports physicians in planning and guiding prostate needle placement using fused MRI and ultrasound images. The goal is to make biopsies and focal therapies more accurate and reproducible, while potentially allowing some procedures to move from hospital operating rooms into ambulatory surgery centres and, eventually, office-based settings.

Biobot says it has deployed more than 80 systems globally and supported over 30,000 patients, with adoption across centres in Europe and Asia-Pacific. The company also points to more than 50 real-world clinical publications supporting its use in targeted transperineal biopsy and precision needle placement.

For a Singapore-based device company, those numbers matter. Medical robotics startups often struggle to bridge the gap between promising engineering and repeatable commercial adoption. Hospitals are cautious buyers, doctors need training, reimbursement can be complex, and regulatory pathways vary widely across markets. The US, while large and lucrative, is also one of the most difficult healthcare markets to enter.

Funding to build evidence and a US beachhead

Biobot plans to use the capital to accelerate international commercialisation, strengthen clinical and economic evidence, and build the commercial infrastructure needed for the next stage of growth. A key part of the plan is a focused US entry strategy.

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

As part of that effort, the company will work with Fogarty Innovation, a California-based medtech accelerator and advisory organisation, to identify priority clinical segments, engage clinicians, and shape its commercialisation approach.

The US market is particularly important because prostate cancer screening, diagnosis, and treatment represent a large and mature specialty segment. At the same time, any new device platform must prove not only that it works clinically, but also that it fits into physician workflows, hospital budgets, and reimbursement systems.

“The objective is not simply to place systems, but to build Mona Lisa into a precision intervention platform for prostate care, connecting imaging, robotic guidance, procedure data and focal treatment workflows,” said Sim Kok Hwee, Deputy Chairman and CEO of Biobot Surgical.

That positioning is important. Biobot is not pitching Mona Lisa merely as a biopsy robot, but as a platform that can sit across the prostate cancer care pathway, from diagnosis to targeted treatment. If it can execute, this could give the company a broader role in clinical decision-making and longitudinal patient management.

Why this matters for Southeast Asia

Southeast Asia has produced relatively few globally visible medical robotics companies, despite rising healthcare demand, ageing populations, and growing investment in specialist care. The region’s healthtech funding has historically leaned towards telemedicine, digital health platforms, insurance technology, and clinic management software. Hardware-heavy medtech, especially robotics, is harder to build because it requires deep clinical validation, manufacturing discipline, regulatory expertise, and long sales cycles.

Singapore is the regional exception. Its combination of public research funding, hospital networks, biomedical manufacturing capabilities, and investors such as Temasek-backed ClavystBio has made it the most likely base for Southeast Asian medtech companies with global ambitions.

ClavystBio was set up by Temasek to help commercialise biotech and medtech ideas from Asia. Its involvement in Biobot signals continued interest in backing companies that can move beyond laboratory innovation into regulated international markets.

“Biobot has the right combination of clinical relevance, international traction, and platform potential,” said Anselm Tan, medtech lead at ClavystBio. “The market is moving toward prostate procedures that are safer, more precise, and more reproducible.”

For the wider region, Biobot’s progress could also have practical implications. Prostate cancer diagnosis and treatment are uneven across Southeast Asia, with gaps in access to specialists, imaging, and advanced intervention technologies. If robotic-guided workflows can lower procedural complexity and support more standardised care, they may eventually help expand access beyond top-tier hospitals, though cost and training remain significant barriers.

The road ahead

Biobot’s next challenge is execution. Its installed base and clinical publications give it a foundation, but scaling in medtech requires more than product adoption at leading centres. The company will need to show that Mona Lisa can deliver measurable clinical and economic value across different healthcare settings.

Also Read: Are biomedicine and healthcare coming of age?

That means convincing hospitals and urologists that the platform improves biopsy accuracy, reduces complications, shortens workflows, or enables new treatment models. It also means building support, training, and service capabilities in markets where device reliability and physician confidence are essential.

The company’s US strategy will be especially critical. Success there could validate the platform globally and strengthen Biobot’s credibility across Europe and Asia Pacific. Failure to gain traction, however, would underline the difficulty of turning Southeast Asian medtech engineering into a global commercial business.

For now, the new funding gives Biobot room to push that ambition. In a region where startup stories are still dominated by software, fintech, and consumer platforms, its progress offers a different narrative: a Singapore-born medical robotics company trying to prove that Southeast Asia can build complex healthcare technology for the world.

The post Singapore’s Biobot Surgical raises US$15.6M to take prostate-care robot global appeared first on e27.

Kazakhstan is not a country most Southeast Asian founders think about when mapping expansion routes. But Magzhan Madiyev, CEO of Astana Hub, wants to change that fast. In seven years, the organisation has helped transform a country with virtually no venture culture into one where 2,200 companies generate US$2 billion annually, tech exports have grown twentyfold to US$1.14 billion, and the nation’s first AI unicorn, Higgsfield, is now valued at over US$3 billion.

The engine behind this shift is not government largesse; it is a calculated infrastructure of talent pipelines, global accelerators, and regulatory architecture built to attract serious founders and investors.

Also Read: How Big Sky Capital and Astana Hub are helping startups scale across Southeast Asia’s technology ecosystem

With Astana Hub opening a presence in Kuala Lumpur, partnerships deepening with Temasek and Quest Ventures, and interest growing from Singaporean investors, Central Eurasia’s most ambitious tech bet is increasingly looking in Southeast Asia’s direction. Here is what Madiyev had to say.

Below are edited excerpts:

Building a startup ecosystem from scratch in an emerging market is notoriously difficult. What was the single biggest structural barrier Kazakhstan had to overcome?

When Astana Hub was conceived, Kazakhstan’s venture market was virtually non-existent. There was no venture culture, no capital, no critical mass of entrepreneurs, and no founder community. The ecosystem was dominated by companies chasing government contracts or running small e-commerce operations.

We had to fight on multiple fronts simultaneously: changing the mindset of specialists, creating innovation-friendly legislation, and overcoming government bureaucracy. None of it was quick.

The results speak for themselves: Almaty has entered the global TOP10 Rising Star ecosystems according to Dealroom’s 2026 index. Astana Hub now hosts over 2,200 companies generating roughly US$2 billion annually, and Kazakhstan’s tech exports surpassed US$1.14 billion in 2025.

Tax incentives can create artificial ecosystems that collapse once the support dries up. How are you ensuring startups are genuinely market-ready?

Every successful tech ecosystem, be it Silicon Valley, Israel, or Singapore, has had government backing at some point. The question is how you deploy it.

We did not ask the state to pour billions into the industry. We asked only for tax incentives and introduced a 1 per cent revenue contribution from participants to reinvest back into the ecosystem. That, combined with full infrastructure, such as education, acceleration, regional access points, and  pathways to Silicon Valley, meant all parties were satisfied without creating dangerous dependency.

The incentives will continue, but they will increasingly push participants to build globally competitive products. If we do not develop our own champions, our economy simply becomes a market for international vendors who are often not even fully taxed here.

How many startups have completed your programmes, and what does the success rate actually look like?

Astana Hub is not a single accelerator; it is an ecosystem of more than 50 programmes serving audiences from school students to institutional investors. Metrics vary accordingly.

Also Read: Big Tech’s efficiency paradox: Record profits, record layoffs

For startup-specific programmes, around 40 per cent of graduates continue building and scaling after completion. Across the broader ecosystem, participant companies have collectively attracted more than US$945 million in investment and generated US$4.9 billion in cumulative revenue over seven years, with a growing footprint in Central Asia, the Middle East, Europe, and North America.

What does the alem.ai Center offer that a startup could not get from Singapore, Dubai, or London?

The alem.ai centre is a vertically integrated AI ecosystem under one roof, something we believe is unique globally. It runs a continuous talent pipeline from children aged 12, through adult re-skilling, to dedicated founder tracks and Big Tech lab placements, all in one building.

That building is also an iconic structure in Central Asia, a deliberate signal from the country’s leadership about where Kazakhstan is placing its long-term bet. In August, we will host part of the International Olympiad in Artificial Intelligence, welcoming national teams from over 60 countries.

Are graduates staying in Kazakhstan or leaving for higher-paying markets?

We do not frame this as brain drain; the world is becoming borderless, and what matters is building a strong community. We actively help founders go abroad to attract investment and launch products in the US, the UAE, and China. Most of them still build their teams and R&D centres back in Kazakhstan.

The most powerful argument for staying is Higgsfield, Kazakhstan’s first AI unicorn, now valued at over US$3 billion, built in Almaty with a team of 300 engineers whose average age is 25. You do not need to leave to build something world-class.

Which international partners have delivered real outcomes, not just signed MOUs?

Several partnerships have moved well beyond paper.

With Google, we launched the Silkway Accelerator — seven batches in, with graduates now carrying a combined valuation exceeding US$500 million.

With OpenAI and Stanford, we ran the AI Leaders programme across nine countries, reaching 800-plus executives.

With Telegram, we are launching the AI Olympiad and ICPC Bootcamp, and the company is opening an AI Lab at alem.ai.

With Draper University and Alchemist Accelerator, Tim Draper has personally invested US$2 million into Kazakhstani startups.

With Apple, a training centre is operating within the ecosystem.

Is there a genuine pipeline of Southeast Asian investors and startups looking at Kazakhstan?

Yes, though the potential has not been fully realised yet. We are opening a presence in Kuala Lumpur, working with Temasek and Quest Ventures, and seeing growing interest from Singaporean investors.

The roadblock is not a lack of opportunity, but a lack of awareness. Southeast Asia is very familiar with its own corridor; Central Eurasia remains undiscovered despite its talent base and fast-growing digital economy. That is precisely what we are working to fix.

Kazakhstan borders both Russia and China. How do you navigate that geopolitically when attracting Western partners?

Kazakhstan is a neutral country by design and we have turned that into a strategic advantage. After 2022, more than 500 global tech companies relocated here from Russia, bringing significant developer talent. From China, we adopt hardware and industrial technology. From the US, we attract capital and infrastructure, including NVIDIA chips. Our most active international hub is in Silicon Valley, where resident companies have attracted over US$300 million in investment.

Also Read: Is the AI industry profitable? Yes, just not where you’re looking

It is a puzzle only a country with Kazakhstan’s geography, neutrality, and foreign policy could assemble.

Venture capital in Central Asia is thin. Are there any homegrown VC funds of meaningful scale?

Kazakhstan leads the venture market across Central Asia and the CIS. The recently established fund of funds, Alem Capital Management, has a first fund of US$100 million, 70 per cent from private capital, with a target of attracting US$1 billion in venture investment over five years. Active local investors include Freedom Holding, MOST Ventures, Big Sky, Astana Hub Ventures, MA7 Ventures, and the Silkroad Angels Club.

What sectors are producing Kazakhstan’s most promising startups, and is it market demand or government money driving them?

AI, edutech, and fintech, and they are driven by talent, not government priorities. That distinction matters. Startups chasing government contracts stay domestic. Kazakhstan’s market is simply not large enough to produce valuations above US$100 million on its own. Unicorns are built for global markets. Higgsfield proves that.

The post Southeast Asia’s investors are sleeping on a US$2B ecosystem next door appeared first on e27.

Thailand-founded enterprise AI company Amity has opened a Singapore office and AI Research & Application Center (ARAC), placing the city-state at the centre of its regional expansion, product development, and planned path to a public listing in 2027.

The move follows Amity’s US$100 million Series D round, led by EDBI, the investment arm operating under SG Growth Capital, alongside Asia Partners and SMDV. Existing and new backers, including CMLIM Capital, also participated.

Also Read: Amity’s US$100M raise signals Southeast Asia’s AI coming of age

Amity said the Singapore hub will serve as its Southeast Asia regional headquarters and the global base for its AI research capabilities. The company is targeting US$200 million in annualised revenue by the end of 2026, after surpassing US$100 million in annualised revenue in 2025. It claims to have grown more than tenfold since 2022.

For Amity, Singapore offers more than a prestige address. It gives the company proximity to regional enterprise buyers, access to AI talent, a recognised regulatory environment, and a launchpad for acquisitions across Southeast Asia and Europe. The company is deploying its latest capital across three areas: accelerating agentic AI development, pursuing strategic M&A, and hiring engineering and commercial talent in Singapore.

Why Singapore matters for Amity

Amity’s bet comes as Southeast Asian enterprises move from AI experimentation to deployment. Banks, telcos, retailers, healthcare groups, logistics players, and large consumer companies are under pressure to automate customer engagement, extract more value from data, and reduce operational complexity across multilingual markets.

This is where Amity wants to position itself. The company’s ARAC will focus on vertical AI models trained on industry-specific data rather than generic datasets. Its current focus areas include agentic AI — systems that can execute end-to-end business processes with limited human intervention — and models tailored for sectors such as retail, telecommunications, and services.

The company argues that Asian enterprises need AI products built for fragmented channels, language diversity, and local business workflows. That is a valid pain point in Southeast Asia, where a single enterprise may operate across markets with different languages, messaging platforms, payment habits, regulations, and consumer behaviours.

“When we started Amity, we believed that Asian enterprises deserved AI that is built for their reality, not tools designed for other markets and retrofitted for ours,” said Korawad Chearavanont, Executive Chairman and Founder of Amity. “Singapore is where that belief comes to life.”

That statement also reveals the company’s broader ambition: to become a regional enterprise AI platform rather than just another SaaS vendor.

The market opportunity

Amity is entering a fast-growing but crowded market. Enterprise AI, customer experience automation, communications analytics, and vertical AI software are attracting capital globally, as companies search for tools that can move beyond chatbots and dashboards into workflow execution.

Global estimates vary, but the enterprise AI market is widely projected to become a multi-hundred-billion-dollar opportunity by the end of the decade. For Southeast Asia specifically, the prize is broader than software revenue alone. A widely cited Kearney estimate has suggested that AI could contribute up to US$1 trillion to Southeast Asia’s economy by 2030.

Also Read: AI user roles surge as Singapore pivots from specialist to mainstream hires

That economic upside is drawing both global players and regional challengers. Microsoft, Google, AWS, Salesforce, OpenAI-linked partners, and a wave of specialised AI startups are all targeting enterprise budgets. Amity’s differentiation will depend on whether it can turn local market knowledge, vertical datasets, and acquisitions into products that large companies are willing to buy at scale.

The company already has some distribution. It claims its platforms serve over 10 million monthly active users across more than 20,000 organisations in over 20 countries. Its portfolio includes Amity Solutions, Tollring, EGG Digital, Amity Accentix, and Amity-Nordstar, covering customer experience, analytics, communications, and voice capabilities.

More recently, Tollring entered into a definitive agreement to acquire UK-based Code Software. If completed, the transaction would expand Amity’s presence across the UK, US, EU, and ANZ markets, especially in communications analytics and recording.

What Southeast Asia could gain

The Singapore hub could bring several benefits to the region if Amity delivers on its plans. First, it could create higher-value AI roles locally. Amity plans to hire across AI research, engineering, and go-to-market functions in Singapore, with up to 60 roles expected over the next three years.

Second, the move could deepen Southeast Asia’s enterprise AI capabilities. Much of the region still depends on imported software stacks designed primarily for Western enterprises. A regional AI player building for Asian languages, regulations, and operating environments could help narrow that gap.

Third, Amity’s Singapore base could strengthen the region’s startup exit and scaling narrative. Southeast Asia has produced strong consumer internet and fintech companies, but fewer enterprise software companies with global ambitions. If Amity reaches its revenue target and lists publicly in 2027, it could become a reference point for the region’s AI and SaaS ecosystem.

Fourth, Amity’s acquisition strategy may create liquidity and growth options for smaller software companies in Southeast Asia. The company has said it will pursue M&A targeting high-potential software businesses in Europe and Southeast Asia. For regional founders building niche B2B software, this could open another route beyond venture funding or slow organic growth.

A tougher phase begins

The challenge is execution. Raising the largest AI/ML round in Southeast Asia for Q1 2026 gives Amity a strong balance sheet, but enterprise AI is not an easy market. Sales cycles are long, integration is complex, and buyers are increasingly sceptical of AI tools that do not produce measurable outcomes.

Agentic AI also carries operational and regulatory risks. Enterprises will need systems that are secure, auditable, and reliable across markets. Singapore’s policy environment may help Amity build credibility here, especially as the country pushes its refreshed National AI Strategy and National AI Council.

EDBI’s backing is also strategically significant. For Singapore, attracting Amity reinforces its ambition to be a regional AI hub not only for multinational labs but also for Southeast Asian technology companies scaling outward.

“Singapore provides a strong foundation for companies looking to develop and scale enterprise AI, with access to deep talent, a trusted and collaborative innovation ecosystem, and strong regional connectivity,” said Yeung Chia Li, Senior Partner at EDBI.

Also Read: Southeast Asia’s AI blind spot is getting bigger

For Amity, Singapore is now the test bed for a bigger question: can a Southeast Asian enterprise AI company build products that compete not only regionally, but globally? If it can, the company’s 2027 IPO plan may be more than a fundraising milestone. It could mark one of the region’s first major enterprise AI scale-up stories.

The post Thailand-founded Amity sets up Singapore AI hub, eyes 2027 IPO appeared first on e27.

There’s a moment in the film Who Am I? where Jackie Chan’s character wakes up with amnesia—no name, no rank, and no recallable past. He’s still capable. Still acting. Still surviving. But the story that once made his actions feel like they belonged to him is gone.

The film keeps circling a single question: Who am I?

But what it stages is something more unsettling. Identity is not something waiting intact beneath the surface. It is something that must be reconstructed when the context that once confirmed it falls away.

Amnesia, in that sense, is not just loss of memory. It is forced exposure to the question of the self.

And increasingly, this is no longer just a cinematic idea.

It is becoming a condition of the age.

Not amnesia, but displacement

The arrival of AI does not erase identity. Memory remains intact. You still know your history, your role, your credentials, and your past achievements. Nothing is removed.

And yet something subtle breaks.

The systems that once stabilised identity—output, expertise, measurable competence—no longer function as reliable mirrors of value. AI can draft cleaner, analyse faster, synthesise broadly, and execute tasks once considered uniquely human markers of capability.

This creates a strange condition: we haven’t forgotten who we are, but we’re losing certainty about what our old markers actually mean. Identity becomes visible, but less anchored. Present, but less confirmed by the world.

It’s not amnesia.

It is a displacement of meaning, an alienation of self.

The identity scaffolding was never just personal

To understand why this feels destabilising, we have to recognise a hard truth: modern identity was never purely innate.

For much of modern history, identity has been socially manufactured through work, a social construct. Thinkers from Marx to Goffman all pointed to the same thread: we don’t just do work. Work tells society who we are—and over time, we inhabit that reflection.

Profession becomes identity. Output becomes a signal. Competence becomes selfhood.

This is the scaffolding AI is now quietly dismantling.

Also Read: What I tell my kids to be able to thrive in the age of AI

When the mirror cracks: Two paths forward

As AI absorbs cognitive labour—coding, writing, analysis, even strategic drafting—the exclusivity of these skills flattens. What once differentiated us becomes abundant.

The loss is double: it’s not just about shifting job descriptions, but about losing a primary anchor of self-worth.

When intelligence is no longer a reliable differentiator, the question changes: What exactly am I expressing when I say “this is what I do”?

When work stops functioning as a stable mirror of selfhood, identity does not disappear. It loses external confirmation.

In moments of systemic disruption, human adaptation rarely moves in one direction. It splits in multiple ways, but generally falls into two dominant forms of responses:

• Compression into optimisation: If intelligence and labour become more machine-readable, humans adapt by becoming highly efficient nodes in a larger wheel. It’s rational. It’s adaptive. But it’s also narrowing.
• Expansion into interpretation: If machines take over execution, what remains human is judgment, framing, and meaning-making. Identity shifts away from output and toward intention: what problems are worth solving, how they’re defined, what gets ignored, and what actually matters.

This isn’t a binary choice. It’s a tension—closer to a Yin–Yang dynamic than a linear progression.

Both emerge at the same time. Neither disappears. The question is not which exists, but which becomes dominant in different contexts and individuals.

AI as a mirror: The self becomes visible

There is a third layer that muddles the id.

AI is not only a displacer of identity. It is also a mirror.

What you get back from AI is shaped by how you think into it—how you frame prompts, what assumptions you carry, what you refine, and what you repeatedly return to. AI doesn’t just extend capability; it reflects cognition.

It reveals the structure of you.

Not who you are in a fixed sense, but how your thinking is organised in real time.

As external validation weakens and internal reflection becomes more visible, identity shifts from something assigned by roles to something observable in patterns of attention.

You begin to see yourself not as a position, but as a way of engaging with the world.

Also Read: Bite-sized innovation: A practical path for SMEs to sustain growth

Liberation through expansion

There is a quieter implication here—one that is easy to miss.

AI not only displace identity structures but also reflects cognitive patterns. It also collapses the barriers between intellectual domains. Plato, poetry, physics, politics, programming—fields that once required years of initiation, institutional access, or rigid disciplinary boundaries—now become fluid, conversational spaces.

• Philosophy: You can question your own self-attachment in dialogue with Zhuangzi: “Now I do not know whether I was then a man dreaming I was a butterfly, or whether I am now a butterfly dreaming I am a man.”
• Physics: You can probe the limits of reality by engaging Niels Bohr on quantum superposition: “Everything we call real is made of things that cannot be regarded as real.”
• Politics: You can confront the fragmentation of social identity through Friedrich Nietzsche—recognising that when AI hyper-personalises your worldview, it constructs a digital tribe of one, isolating identity from the native cultural fabric.
• Programming: You can visualise recursive self-awareness through a simple Python loop—an architecture that mirrors how identity continuously reflects and redefines itself:

def identity(input_self):
    # AI mirrors human thought, which mirrors AI output
    reflection = f"AI reflects: {input_self}"
    print(reflection)
    return identity(reflection)  # The endless loop of self-definition

• Poetry: You can interrogate your own performativity by stepping onto Shakespeare’s stage: “All the world’s a stage, / And all the men and women merely players.” As AI automates the script of daily labour, we are no longer confined to being “merely players” in predefined roles—forcing a more difficult question: who are you when the performance falls away? If these disciplines ignite curiosity, the exploration does not stop there.
• Psychology: You can turn inward with Carl Jung: “Until you make the unconscious conscious, it will direct your life, and you will call it fate.” Because the cost of entry has collapsed, the outcome is not simply frictionless access to information. It is access to entirely new modes of thinking. In this sense, the world does not just expand. It becomes exactly as large as your willingness and ability to move across it.

Deconstructing the shell, eeconstructing the self

This creates a quieter awakening.

When identity is no longer defined primarily by output, it becomes harder to outsource selfhood to systems of performance. What remains is not absence, but exposure.

Exposure to how attention moves. How curiosity unfolds. How judgment forms. How questions are shaped—and reshaped.

Identity begins to shift—from what you produce to how you engage.

This is not comfortable.

But it is clarifying.

Because it reveals something long obscured by the apparent stability of roles: identity was never something we simply possessed. It was something continuously negotiated through systems that reflected us back to ourselves.

This is where the self begins to awaken—and that is larger than AI.

In other words, identity does not emerge from static labels, but from dynamic interaction—from the ongoing “ping” between self and world.

Not a title. Not a role. But a pattern of engagement with the universe itself.

Also Read: Workers sprint ahead of bosses in AI adoption in Singapore, exposing a transformation gap

The one question AI cannot answer

There is a paradox at the centre of all this.

AI can simulate reasoning. It can generate language. It can approximate styles, arguments, and even forms of creativity.

But there is one question it cannot answer for you.

Who are you?

Not as a profile. Not as a dataset. Not as an aggregation of outputs.

If work no longer defines identity, and intelligence is no longer uniquely human, then “Who am I?” stops being a philosophical abstraction.

Cogito, ergo sum?

The only role you cannot outsource

In the end, identity becomes something like a film that no Generative AI can recreate.

A narrative without a pre-trained model. A story without a dataset.

You are not the prompt. You are not the output. You are the one who must live the sequence.

You are the main actor. You are the scriptwriter. You are the only director across every scene.

Ultimately, “Who am I?” means synthesising your own humanness together, frame by frame.

—

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 Who am I in the age of AI? Identity, displacement, and awakening appeared first on e27.

When I was 12, I walked around my uncle’s warehouse in the Netherlands. He’d started a jewellery import business and grown it into a big business. I stood there thinking: he started with nothing and built all of this, that’s what I want to do!

I also noticed the Porsche. And the house. I wanted those too. (I gave up on the Porsche a while ago.)

That wasn’t a career decision. I didn’t weigh the pros and cons of entrepreneurship versus getting a job. I didn’t have the vocabulary for any of that. Something clicked, standing in that warehouse, and it never unclicked.

I meet people every week who want to start a company. They’ve done the research. They’ve read the books. Some of them have compared the risk profile of founding versus staying in their corporate job. A few have actual spreadsheets.

If you need a spreadsheet to decide whether to start a company, you probably shouldn’t start one.

I know how that sounds. But I’ve watched people try this for 25 years, and the pattern holds. The founders who survive the first two years (the ones still standing when the money runs out, the co-founder leaves, the product doesn’t work and needs to be rebuilt from nothing) almost never chose this the way you choose an MBA program. They chose it the way you choose breathing. They couldn’t not do it.

Also Read: The founder’s labyrinth: Why the US$2T climate finance industry is failing ‘atoms’ in SEA

Right now, founding looks like a career option. AI tools let one person build a product in a weekend. Capital is around. The stories are everywhere. So people jump. They leave a corporate job, register a company, start showing up at startup events. For a while, it feels like being a founder.

Then the other stuff starts. The financial spreadsheets you hate making. The contracts you can’t afford to get wrong. The client who’s unhappy and threatens to sue you. The 19th pitch that doesn’t land. The employee who wants more salary and 20 days off.

Corporate life trains you to be a specialist. You get good at one thing, inside a structure someone else built. Founding is the opposite. You do everything nobody else wants to do, for as long as it takes, with no certainty any of it will work. Most people from corporates come with the exact opposite preparation for this.

That’s where the itch matters. When all of it hits you at once (and it will), the only thing keeping you in the chair is that you can’t imagine sitting anywhere else.

If you’re weighing whether to start a company the same way you’d weigh a job offer, you’re in the wrong frame. The people who build things that last didn’t weigh it. They just started. Usually before they were ready. Usually before anyone around them thought it was a good idea. And they kept going, no matter what.

The post If you need a spreadsheet to decide whether to start a company, you probably shouldn’t start one appeared first on e27.

Modern life has become deeply optimised.

There are apps to improve our focus, watches to track our sleep, systems to organise our mornings and endless advice on how to squeeze more output into the same 24 hours. Social media feeds are filled with productivity routines, side hustle culture and carefully engineered lifestyles designed to maximise performance.

And yet, many people feel more emotionally exhausted than ever.

Not necessarily because they lack ambition, but because optimisation itself has quietly become a full-time mindset.

For years, modern work culture has operated on the assumption that the solution to overwhelm is better efficiency. Better systems. Better routines. Better time management. More automation. More hacks.

But increasingly, many people are no longer looking for ways to do more.

They are looking for ways to live that actually feel sustainable.

Productivity culture expanded beyond work

What began as workplace optimisation slowly spread into every corner of life.

Careers are no longer enough on their own. People are encouraged to build personal brands, monetise hobbies, maintain online visibility and continuously improve themselves professionally and personally. Even activities that were once considered leisure now come with subtle pressure to become productive.

Exercise becomes performance tracking. Reading becomes self-improvement. Vacations become content opportunities. Social media becomes networking. Hobbies become side hustles.

Life itself starts to feel operationalised.

The result is that many people no longer feel fully “off”, even during their downtime. Notifications continue. Messages continue. The mental tabs remain open.

The rise of always-on work culture has also blurred the boundaries between productivity and recovery, contributing to rising levels of what many professionals now describe as digital burnout.

At some point, the issue stops being about workload alone. It becomes about the inability to psychologically disengage.

Also Read: The AI productivity gurus are bluffing too

The problem is not ambition, it is an unsustainable ambition

This distinction matters.

Most people still want meaningful careers, financial stability (or freedom, which explains the popularity of creating passive income streams) and opportunities to grow. Founders still want to build successful businesses. Professionals still want purpose and progress.

The issue is not that people suddenly want less from life.

The issue is that many modern systems reward constant optimisation without acknowledging human limits.

In many industries today, being busy has become intertwined with being valuable. People are expected to move quickly, stay visible, adapt constantly and remain mentally available at all times. Even rest is often framed as recovery for more productivity later.

But human beings are not designed to remain perpetually “on”.

Even high performers eventually experience the effects of fragmented attention, continuous responsiveness and prolonged mental stimulation. And unlike traditional burnout, modern exhaustion is often quieter and more difficult to identify because it accumulates gradually.

For many professionals, the issue is no longer just long hours, but prolonged exposure to fragmented attention, constant responsiveness and elevated stress hormones throughout the day.

The irony is that many modern workers are not necessarily lacking productivity tools. They are lacking meaningful opportunities for psychological recovery.

People are not just seeking rest, they are seeking permission to be “off”

One of the most overlooked aspects of modern productivity culture is the guilt people increasingly feel around doing nothing.

There is now subtle pressure to optimise almost every waking hour. If someone is resting, they should be resting productively. If they are scrolling social media, it should somehow lead to inspiration, learning or monetisation. Even hobbies increasingly come with pressure to become content, side income or personal branding opportunities.

But increasingly, what many people actually want is far simpler.

They want time to:

• Spend time with family without multitasking
• Enjoy hobbies whenever they want
• Rest without guilt
• Be mentally unreachable for a while
• Experience moments that are not constantly interrupted by notifications, deadlines or content demands

In other words, they want enough emotional and mental space left to actually enjoy the lives they are working so hard to build.

And that desire is not laziness. It is a response to years of overstimulation and perpetual optimisation.

Sustainable living is not about doing less, it is about designing better priorities

The answer is probably not another productivity framework.

Nor is it abandoning ambition altogether.

If anything, sustainable ambition may become one of the most important skills modern professionals and founders need to develop.

Also Read: The productivity pivot the Philippines can’t delay

That means recognising:

• Not every opportunity deserves a yes
• Not every platform deserves constant attention
• Not every hobby needs monetising
• Not every hour needs to be maximised
• Not every moment of rest needs justification

It also means organisations may need to rethink what sustainable performance actually looks like.

Many companies still reward responsiveness over deep thinking, visibility over focus and busyness over meaningful outcomes. But constant interruption and overstimulation eventually reduce creativity, emotional resilience and long-term decision-making quality.

The businesses that adapt best in the future may not simply be the ones moving fastest.

They may be the ones capable of building cultures where people can sustain high-quality thinking and meaningful work over time without permanently operating in survival mode.

Because people are not searching for another productivity hack but are searching for lives that still feel emotionally livable.

Perhaps the real luxury now is spaciousness

For years, success has often been associated with acceleration: faster growth, faster scaling, faster output, faster responses.

But perhaps the next real luxury is not speed.

Perhaps it is spaciousness.

The ability to think slowly sometimes. To be unreachable occasionally. To spend time with people you love without simultaneously checking notifications or replying to messages. To enjoy moments that are not being turned into content, strategy or productivity metrics.

Technology will continue evolving. Work will continue changing. Economic pressures are unlikely to disappear anytime soon.

But eventually, both individuals and companies may need to ask a more uncomfortable question:

What is the point of building successful lives if we are too mentally exhausted to actually experience them?

Because perhaps the future advantage will not belong to the people who can optimise themselves endlessly.

It may belong to those who can build ways of working and living that remain psychologically sustainable over the years, not just being productive for quarters.

—

The post People don’t want productivity hacks anymore, they want sustainable ways to live appeared first on e27.

Southeast Asia has never followed a single digital playbook. A customer in Thailand may expect to interact with a brand through LINE. A shopper in Indonesia or Malaysia may prefer WhatsApp. In Vietnam, Zalo remains deeply embedded in daily communication. In the Philippines, Messenger continues to shape how people connect, discover, and transact.

This makes the region different from many Western markets, where customer journeys are often designed around websites, email, apps, and scheduled support hours. In Southeast Asia, the customer journey is increasingly conversational, mobile-first, and always on.

That is why AI agents will not simply become another customer service tool. They will reshape how brands design the entire customer journey, from discovery and onboarding to service, retention, and reactivation.

Southeast Asia is already a messaging-first region

The case for AI agents starts with user behaviour.

Southeast Asia has high levels of internet and social media adoption. We Are Social’s Digital 2025 Singapore report notes that internet adoption across Southeast Asia reached 78.2 percent, while social media use stood at 61.5 percent of the total population. In Singapore, 92.4 percent of internet users are active on social media.

Country-level data shows how deeply digital behavior is embedded across the region. In Thailand DataReportal found that there were 65.4 million internet users at the start of 2025, with internet penetration at 91.2 percent. The country also had 51 million social media user identities, equal to 71.1 percent of the population. LINE reported 56 million monthly active users in Thailand, equivalent to 78.2 percent of the total population and 85.7 percent of internet users.

In Vietnam, DataReportal recorded 79.8 million internet users and 76.2 million social media user identities in January 2025. In the Philippines, there were 97.5 million internet users and 90.8 million social media user identities at the start of 2025. Singapore, meanwhile, had 5.61 million internet users and 5.16 million social media user identities, equal to 95.8 percent and 88.2 percent of the population respectively.

These numbers point to a simple reality: brands in Southeast Asia are not trying to bring customers online. Customers are already online. The harder challenge is meeting them in the channels where they already spend time, in the language they prefer, and at the moment they need help.

Also Read: The new cybersecurity threat: Why AI agents are the wild card in enterprise security

Customers now expect always-on engagement

The traditional customer journey assumes a certain rhythm. A customer sees an ad, visits a website, submits a form, receives an email, waits for a reply, and eventually speaks to a salesperson or support agent.

That journey is becoming too slow for Southeast Asia’s mobile-first consumers.

In messaging-first markets, customers often expect brands to behave more like people in their contact list. They want to ask a question, get a response, clarify a detail, change a booking, check delivery status, or complete a transaction without switching channels. If a brand takes hours to respond, the customer can easily move to another seller, another platform, or another app.

This is where AI agents change the equation.

Unlike traditional chatbots, which are usually limited to fixed menus and scripted answers, AI agents can understand intent, retrieve context, take action, and escalate when needed. They can support customers outside office hours, handle repetitive questions, personalise recommendations, and help human teams focus on more complex or sensitive interactions.

Globally, companies are already moving in this direction. Zendesk’s 2025 CX Trends report found that consumers increasingly expect AI interactions to feel more human, personalised, and engaging. The report also describes a widening gap between companies that embrace AI in customer experience and those that remain tied to traditional support models.

For Southeast Asia, the opportunity is even more urgent because customer journeys are fragmented across countries, languages, channels, and behaviours.

Local behaviour matters more than global templates

One mistake brands often make in Southeast Asia is assuming that a customer engagement strategy built for the US or Europe can simply be localised with translation.

But localisation is not only about language. It is also about behaviour.

A customer in Bangkok may be comfortable using LINE for brand updates, payments, service reminders, and support. A customer in Jakarta may discover a product through social content, ask questions through WhatsApp, and expect the conversation to continue with a human seller. A customer in Ho Chi Minh City may use local platforms as part of their daily routine in ways that do not map neatly to Western customer journey models.

This means brands need AI agents that understand context, not just words. They need to know when to be proactive, when to wait, when to escalate, and when a conversation requires local nuance.

For example, an AI agent for a bank in Southeast Asia should not only answer questions about loan eligibility. It should be able to guide a customer through documentation, remind them of missing steps, hand off to a human agent when trust is needed, and operate across local languages and channels.

For e-commerce, an AI agent should not only track orders. It should help customers compare products, ask preference-based questions, recover abandoned carts, handle delivery issues, and continue the conversation after the purchase.

The winning brands will be those that design AI agents around local journeys rather than forcing customers into imported workflows.

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

AI agents can connect fragmented customer journeys

Southeast Asia’s digital economy is full of fragmented touchpoints. Customers move between ads, marketplaces, super apps, social platforms, messaging apps, call centres, and offline interactions. For businesses, this creates a major challenge: the customer journey is often distributed across systems that do not talk to one another.

AI agents can become the connective layer.

When integrated properly, an AI agent can recognise a returning customer, understand past interactions, continue a conversation across channels, and recommend the next best action. This moves customer engagement from reactive support to proactive journey orchestration.

This is especially important in Southeast Asia, where businesses often operate across multiple countries with different languages, channels, and service expectations. Agora’s 2025 partnership with WIZ.AI, for example, focused on enterprise-ready AI agent solutions with multilingual support and contextual understanding for call centres.

The broader shift is also being recognised by global consulting firms. BCG argues that AI-powered agents will enable brands to deliver more personal customer interactions at lower cost-to-serve, making customer experience less tedious for consumers and more efficient for businesses.

Human agents will still matter, but their role will change

The rise of AI agents does not mean human teams will disappear. In Southeast Asia, where trust, empathy, and relationship-building remain important, human support will continue to matter.

What will change is the role of human agents.

Instead of spending most of their time answering repetitive questions, human teams can focus on high-value conversations: complex complaints, sensitive financial decisions, healthcare concerns, enterprise sales, VIP customers, or moments where emotional intelligence is needed.

AI agents can handle the first layer of engagement, collect information, summarise context, and route the customer to the right human expert. This makes the handoff faster and more informed.

For customers, the experience becomes smoother. They no longer need to repeat the same issue multiple times. For businesses, teams can scale support without sacrificing quality.

The next customer journey will be conversational

In Southeast Asia, AI agents will reshape customer journeys not because the technology is new, but because it fits how consumers already behave.

The region’s customers are mobile-first, messaging-first, and increasingly unwilling to wait for support that follows office hours or rigid workflows. For brands, this creates a clear opportunity: use AI agents not as a chatbot upgrade, but as the connective layer between discovery, service, sales, and retention.

The companies that win will be those that build around local behaviour. In Southeast Asia, a better customer experience will come from conversations that are instant, contextual, multilingual, and easy to continue.

—

The post Why AI agents will reshape customer journeys in Southeast Asia appeared first on e27.

The product labelled MVP is usually not a minimum. It is a compromise between anxiety, ambition, internal politics, and the hope that one release can settle more questions than it ever will. By the time it reaches execution, the so-called minimum version has absorbed edge cases, stakeholder comfort features, reporting needs, admin controls, future-proofing logic, and enough polish to make the team feel less exposed when customers finally see it.

What gets shipped is not a minimum. It is a fear-managed version of the idea.

Why teams keep getting this wrong

Most teams do not overscope because they are careless. They overscope because their definition of minimum is quietly corrupted.

Product thinks minimum should still feel strategically credible. Engineering thinks that the minimum should not create avoidable technical debt. Design thinks minimum should not feel incomplete to users. Sales thinks the minimum should not be embarrassing in front of prospects. Leadership thinks minimum should still look meaningful enough to justify the bet.

Each of those instincts is understandable. Together, they are how small ideas become large commitments.

The core problem is that minimum gets interpreted through internal discomfort rather than market need. Teams are not asking, “What is the smallest thing that can teach us whether this matters?” They are asking, “What is the smallest thing we can ship without feeling exposed?”

Those are very different questions.

The first creates learning. The second creates bulk.

Minimum is not about feature count

One reason the MVP discussion gets so muddled is that people talk about it as a scope exercise. They reduce the challenge to cutting screens, dropping workflows, or trimming integrations. That is part of the work, but not the heart of it.

A real minimum is not defined by how little you build. It is defined by what must be true for the test to mean something.

Also Read: How a cross-border tech team built a fintech MVP in 3 months

That means the minimum should be tied to evidence, not volume. If your product idea depends on customers trusting the output, then credibility is part of the minimum. If your concept relies on repeated use, then enough continuity for a second use matters more than broad functionality. If the whole point is to prove willingness to adopt, then the minimum may sit less in the interface and more in whether the user can actually get to value without excessive explanation.

This is where many startup teams lose discipline. They cut obvious features while keeping hidden complexity. They remove visible scope but preserve all the machinery underneath it. They tell themselves the product is lean because the roadmap looks shorter, even though the build still assumes full workflow coherence, broad edge case support, and an operational model fit for a much more mature product.

The result is a product that looks smaller on paper but behaves like a much bigger bet.

What real minimums actually look like

A useful way to think about minimum is to stop treating it as one thing. In practice, there are several minimums that matter, and confusing them is how teams get into trouble.

The first is the minimum value. What is the smallest meaningful improvement in the user’s world that they would actually notice and care about? Not admired in a demo. Not politely praised in feedback. Actually care about enough to change behaviour.

The second is the minimum proof. What is the least you need to observe to know whether the problem is real, the proposition is resonating, or the workflow has legs? This is often much smaller than the team wants to believe. Most early products do not fail because they lacked feature breadth. They fail because nobody got honest about what evidence would count as real progress.

The third is the minimum credibility. This is where product belief often becomes unhelpful. Some ideas can survive with a rough edge. Others cannot. If you are asking a user to trust a recommendation, a financial action, a workflow decision, or something that touches their customers, quality and coherence may be part of the minimum from day one. Not because you are polishing for vanity, but because, without credibility, the test itself becomes false.

The fourth is the minimum operability. Can the thing actually be supported, explained, monitored, and recovered when it breaks? Startups often ignore this because it feels too early. Then they wonder why the product produces noisy feedback that is impossible to interpret. If usage fails because onboarding is confusing, support is absent, or obvious issues cannot be diagnosed, you are not testing the product cleanly. You are testing a muddled experience.

Real minimums sit at the intersection of those four questions. Anything beyond them deserves much more suspicion than most teams apply.

The hidden reason MVPs grow

There is another force at work here, and product leaders need to name it more honestly. Large MVPs are often a way of buying emotional reassurance.

A bigger first release lets more people feel covered. It reduces the number of awkward questions before launch. It creates the impression of momentum. It allows teams to believe that if adoption is weak, the issue must be go-to-market execution rather than the shape of the product itself.

In other words, size becomes a defence against ambiguity.

Also Read: Founders, stop listening to mentors who tell you to build an MVP

When an MVP is too big, you are not managing risk; you are relocating it

The usual argument for a broader MVP is risk reduction. Teams say they need more before launch because they want to avoid customer disappointment, reduce rework, or make the proposition more complete.

Sometimes that is valid. Often it is a sleight of hand.

What they are really doing is shifting risk from the market to the build. Instead of risking that customers might not engage with a thinner offer, they risk extra months of effort, deeper architectural commitment, noisier prioritisation, and greater internal attachment to the solution. The commercial uncertainty has not disappeared. It has simply been wrapped inside a larger delivery motion.

That is a dangerous trade because it creates the illusion of progress while increasing the cost of being wrong.

The better question is not “what can we cut?”

Most teams approach MVP scoping in the wrong direction. They start with the full imagined product, then ask what can be removed. That approach almost always leaves too much intact because the emotional centre of gravity remains with the larger vision.

A better way is to start with the evidence you need and work forward from there.

• What are we trying to learn?
• What user behaviour would count as real traction?
• What has to exist for that behaviour to happen credibly?
• What can fail quietly without invalidating the test?
• What is the narrowest path to value we can support properly?

These questions change the conversation. They force the team to design for proof rather than aspiration. They also make it easier to identify fake necessities, which are features that sound important only because the team has already become attached to the more complete story.

This is not just a scoping technique. It is a discipline of strategic honesty.

—

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 How to build a real MVP: Start with evidence, not features appeared first on e27.