The cryptocurrency sector is moving higher in a coordinated fashion. Bitcoin rose 1.70 per cent to US$85,002.11 over the past 24 hours. The total crypto market climbed 1.16 per cent to US$2.89T. This advance rests on two powerful forces. Institutional optimism and policy progress are working together. The result is a rally that looks more grounded than a purely speculative surge. Digital assets are responding to concrete developments that could reshape how traditional finance interacts with this space.

The primary catalyst arrived on October 1. Citigroup analyst Alex Saunders raised the bank’s base-case Bitcoin forecast to US$113,000 from US$82,000. Saunders cited renewed ETF inflows and positive sentiment. At the same time, the SEC proposed a tailored custody framework for investment advisers. This framework would allow self-custody under certain conditions.

The proposal addresses a key operational hurdle that has limited professional product offerings. These two events signal growing Wall Street validation and clearer oversight pathways. They directly boost large-player buying interest. Traders read these developments as evidence that barriers to broader professional participation are slowly coming down.

Market structure data supports the view that this advance is orderly rather than frenzied. Derivatives open interest increased 6.28 per cent. Liquidations fell sharply by 61.8 per cent. That combination points to a controlled move without a leverage flush. Bitcoin dominance held at 58.96 per cent. This stability suggests core assets led the rally rather than speculative altcoins.

Social sentiment for Bitcoin remained mildly bullish with a net score of 5.26. Broader social sentiment registered a net score of 5.85. The Fear & Greed Index reading of 68 places the sector in Greed territory. The top-trending narrative is the US Strategic Crypto Reserve, which gained 1.41 per cent over the same period. These metrics paint a picture of steady capital inflow. They do not show forced selling pressure or excessive speculation.

Also Read: Uptober or downtober: Will Bitcoin’s 19% seasonal average survive US$100 oil?

The correlation between this asset class and traditional risk assets deserves attention. Crypto shows an 80 per cent correlation with the Russell 2000 ETF (IWM). This relationship suggests that digital assets increasingly trade in tandem with broader risk sentiment. They no longer behave as an isolated asset class.

The implication is significant. Crypto participants must now weigh macroeconomic factors alongside sector-specific developments. The upcoming US jobs report on October 2 looms as an immediate trigger. That report could influence rate expectations and risk appetite across both traditional and digital markets.

From a technical perspective, Bitcoin faces a substantial sell wall between US$85,000 and US$86,500 on Binance. Traders noted this wall on September 24. The 365-day moving average near US$80,000 provides major support. If Bitcoin holds above US$83,000, it could challenge the key resistance zone between US$85,000 and US$86,500. A drop below US$82,000 risks a pullback toward the US$80,000 support. The path of least resistance remains upward.

Progress hinges on absorbing the overhead supply. Whether spot volume expands on a decisive break above US$85,000 will confirm genuine buying interest over passive resistance. The key question for the immediate future is whether Bitcoin can close above US$85,500 within the next 48 hours. Such a finish would signal a convincing breakout from its recent range.

For the broader crypto sector, the technical picture is equally instructive. The market cap is testing the 23.6 per cent Fibonacci retracement level at US$2.85T. Near-term resistance sits at the recent swing high of US$2.94T. A hold above US$2.85T suggests traders are pricing in the bullish catalyst. A failure here could see a retest of the 50 per cent level at US$2.76T.

If the SEC’s proposal maintains positive momentum, a test of the US$2.94T ceiling is likely. A drop below US$2.85T could signal a return to consolidation. The US$2.85T to US$2.94T range represents the immediate battleground. A close above it could signal a continuation toward US$3.03T.

Also Read: The sovereign shift: Why nation states are trading gold for Bitcoin

Spot Bitcoin ETF AUM now stands at US$111.13B. That figure rose from US$110.97B yesterday. This modest but consistent increase aligns with the inflow trend cited by Citigroup. Sustained weekly ETF flow data will be crucial to confirm whether professional capital continues to enter the space at a meaningful pace.

The 60-day public comment period on the SEC’s custody proposal will also bear watching. Formal responses from major financial institutions could shape the final oversight landscape. Any softening of large-player backing or negative policy surprises could quickly alter the current trajectory.

My point of view is that this rally rests on more durable foundations than previous speculative surges. The combination of a major bank’s endorsement and proactive policy clarity provides a firm foundation for Bitcoin’s latest leg higher. The trading environment is not simply responding to hype. It is responding to concrete shifts in the institutional and policy landscape.

The alignment between Bitcoin’s strength and the broader market’s uptick suggests a healthy dynamic in which core assets lead rather than lag. This sector remains vulnerable to macroeconomic shocks, particularly from the upcoming jobs report. The 80 per cent correlation with the Russell 2000 ETF suggests that any risk-off sentiment in traditional markets could quickly spill over into crypto.

The near-term outlook is cautiously optimistic. A tangible reduction in policy uncertainty anchors the uptick. That reduction could unlock new professional capital. Positive sentiment and Bitcoin’s leadership provide a supportive backdrop. The key question for the week is whether the sector can convert this policy goodwill into a sustained breakout above US$2.94T.

For Bitcoin specifically, the challenge is whether it can absorb the sell wall at US$85,000 to US$86,500 and establish a new range. If these technical barriers fall, the path toward US$113,000 as forecast by Citigroup becomes more plausible. If they hold, the market may need to consolidate further before attempting another leg higher. The next 48 hours will be critical in determining which scenario unfolds.

—

The post The October 2 US jobs report lands soon. Could it make or break Bitcoin’s rally? appeared first on e27.

For years, companies have treated Google rankings as the front door to online discovery. A new study from Manila-based digital marketing agency Spiralytics suggests that door is no longer the only one that matters.

As consumers increasingly ask ChatGPT, Google Gemini and Perplexity for product recommendations, comparisons and buying advice, brands face a new visibility test: whether artificial intelligence systems mention them at all.

In the Philippines, the early results look encouraging on the surface. Spiralytics found that Philippine brands appeared in 83.2 per cent of the AI answers where they were deemed relevant, compared with 40 per cent for a comparison group of US brands.

Also Read: The Philippines does not need to build AI to have an AI advantage

But the more important finding may be who is doing the talking. According to the study, 95 per cent of the sources cited by AI engines when discussing Philippine products and services came from third parties rather than the brands’ own websites. In practice, that means news articles, online directories, industry roundups and other external references are helping decide how companies appear in AI-generated recommendations.

For Southeast Asian businesses that have spent years optimising websites for search engines, the study points to a shift in how online trust is being assembled. It is no longer enough to publish a clean product page or rank for a keyword. AI systems appear to be piecing together brand identity from a wider set of signals, including whether credible external sources can confirm what the company does.

A new scoreboard for online visibility

Spiralytics conducted the study between 4 June and 3 August 2026. It generated 1,146 real buyer questions for 26 brands, screened the questions for relevance, and evaluated responses across ChatGPT, Google Gemini and Perplexity. In total, the agency reviewed and stored 1,911 AI answers.

The study frames AI visibility as a second scoreboard alongside traditional search engine optimisation (SEO), the practice of improving a website so that it ranks higher on search engines such as Google. AI visibility is slightly different. It asks whether a brand is named, recommended or correctly understood when a consumer turns to an AI assistant for advice.

That distinction matters because Southeast Asia’s internet economy is highly fragmented. Consumers often move between search, social media, marketplaces, chat apps and review sites before making a purchase. In markets such as the Philippines, Indonesia and Vietnam, where mobile-first behaviour is the norm, conversational AI could become another layer in that discovery journey rather than a replacement for search.

The risk for brands is that AI answers can feel authoritative even when they are incomplete. If a company is missing from a relevant recommendation, misidentified, or described through outdated third-party material, the consumer may never visit its website to verify the details.

Third parties hold the pen

One of the study’s sharper findings is that Philippine brands are visible, but not necessarily in control of their visibility. While they appeared frequently in AI answers, their own websites were rarely the cited foundation for those responses.

This should concern marketing and communications teams. Corporate websites remain useful for explaining products, pricing, leadership and services. But if AI engines are leaning heavily on external sources, then earned media, directories, analyst-style lists and independent explainers become more influential in shaping how a company is represented.

“Surprisingly, backlink authority did not predict AI visibility anywhere in our data,” said Gabe Tagala, Head of SEO at Spiralytics. “It does not mean SEO is obsolete. Links still matter where rankings live. But AI keeps a different score: whether the engine can tell which company you are, and if reputable third parties can vouch for you.”

That point is especially relevant in Southeast Asia, where many companies operate through holding structures, subsidiaries, regional brands and local-language identities. A business may be well known offline or on social platforms, but still poorly defined in the open web sources that AI models use to generate answers.

Gemini led on Philippine queries

The study also found meaningful differences between AI engines. For Philippine queries, Google Gemini had the highest brand mention rate at 62.6 per cent, followed by Perplexity at 57.4 per cent and ChatGPT at 45.2 per cent.

Spiralytics described the engines as behaving like “different editors”. That is a useful way to think about the problem. Each tool may weigh sources, freshness, citations and confidence differently. A brand that appears in one engine’s answer may be absent in another.

Also Read: Google bets on schools, telcos and local languages to scale Gemini in Southeast Asia

The study found that absence can also persist. Philippine brands were missing from the three engines’ responses for one in six commercial questions where they should have appeared. In repeat testing on one brand across 11 identical runs, 90 per cent of unanswered questions stayed unanswered each time. Spiralytics said this suggests omission may continue until a brand’s identity signals or third-party coverage change.

For companies, this creates a new monitoring challenge. Tracking search rankings alone will not show whether AI systems are recommending a competitor instead, or whether they are failing to mention the brand in a category where it has a legitimate claim.

The local naming problem

The research also highlights a familiar problem in emerging markets: names do not always travel cleanly across databases and AI systems. Spiralytics found that engines sometimes confused Philippine entities with overseas namesakes, declined to identify a company clearly, or folded subsidiaries into parent conglomerates.

This is not a minor technical issue. In sectors such as fintech, healthcare, education and logistics, a mistaken identity can damage trust or divert demand. For regional companies expanding across Southeast Asia, the challenge compounds as brands adapt to different languages, regulators and market conventions.

The comparison with the US market is revealing. In the US sample, competitor websites accounted for 60.7 per cent of AI citations. In Philippine answers, competitor websites accounted for 36 per cent. That leaves more room for editorial and directory sources to shape AI responses in the Philippines, at least for now.

Jimmy Cassells, co-founder and CEO of Spiralytics, argued that this gap may close quickly. “This window won’t stay open,” he said. “Brands who will move now and treat SERP and AI visibility as a connected strategy are the ones that will still be recommended a year from now.”

Also Read: The AI marketing backlash story doesn’t actually fit Southeast Asia

The broader lesson for founders and operators is simple. AI discovery is becoming part of brand building. Companies need clear websites, but they also need consistent third-party validation and unambiguous public information about who they are, what they sell and where they operate.

For Southeast Asian startups, this is both a threat and an opening. Larger incumbents may have bigger marketing budgets, but younger companies can still earn visibility if they are clearly described across credible public sources. In an AI-mediated internet, the brands that are easiest to understand may become the easiest to recommend.

The post Why Philippine companies are winning AI mentions but losing control of context appeared first on e27.

Jonathan Honig, who owns about 9% of MoneyHero’s Class A shares, wants the Nasdaq-listed comparison platform sold. His open letter asks the board to hire an independent adviser and explore a sale before a 5 October deadline.

Honig’s complaints are hard to dismiss: no permanent CEO six months after Rohith Murthy’s exit, revenue down from US$80.7 million in FY2023 to US$73.4 million in FY2025, and a share price more than 88% below its 2023 debut. At US$0.675 a share, the market values SingSaver, Seedly, Moneymax and 10.1 million members at roughly US$30 million, barely above the company’s US$28.2 million cash pile. Honig nearly doubled his stake as the stock fell, betting the cash sets a floor. His proposed US$1.50 a share would more than double his money.

The arithmetic of control tells a different story. Richard Li’s sponsor entity holds 38.3% of the equity and 81.1% of the votes through super-voting Class B shares. Honig can embarrass the board, but he cannot outvote it. A pressured sale also carries risks: a take-under priced off cash, a break-up that thins Southeast Asia’s comparison market, or a buyout by the controller itself. PropertyGuru’s 52% premium shows exits are possible, but MoneyHero lacks its pricing power.

REGIONAL

PSA, Granite Asia launch US$50M fund for supply-chain tech: The G&P Strategic Innovation Fund will co-invest alongside Granite Asia’s existing funds, backing port, logistics and supply-chain startups that can move beyond pilots. PSA, which handled over 105 million TEUs in 2025, offers deployment sites across 45 countries.

Amazon pays US$75M for Selangor land to build a data centre: Amazon Data Services Malaysia is buying a 184,000 sq m freehold plot in Dengkil, Sepang, from developer Sunsuria. The deal needs state and data-centre task force approvals, with completion due by Q2 2027.

Zeya Health buys ConnectLah to automate clinic bookings with AI: Antler-backed Zeya absorbs ConnectLah’s WhatsApp-based booking workflows after posting 15-fold revenue growth. It serves providers in six markets and targets 1,000 by mid-2027. ConnectLah founder Jules Pereira-Gomes joins as strategic advisor.

MoneyHero shareholder urges board to explore sale after stock slump: Jonathan Honig, who owns about 9% of MoneyHero’s Class A shares, wants an independent adviser hired as the Singapore-based platform battles leadership uncertainty, stalled revenue growth and a sharp fall in its Nasdaq-listed shares.

MDEC chief Anuar Fariz Fadzil to exit after US$44B investment push: Anuar leaves Malaysia’s national digital-economy agency when his contract ends on 2 October, closing a two-year tenure that sharpened MDEC’s focus on AI, high-value investment and measurable economic outcomes.

Temasek to open first Middle East offices in Riyadh and Abu Dhabi: Due to open in 1H 2027, the new hubs will widen Temasek’s deal access into Central Asia and Africa and lift its network to 15 offices in 11 countries. The state investor manages a S$518B net portfolio.

Malaysia’s top 100 startups generated US$2.98B in FY2024 revenue: Cradle’s inaugural Top 100 list shows an 87% average CAGR, with 36 startups earning over MYR1M in net profit and 58 operating abroad. AI and data-driven businesses make up 66% of the cohort.

Kawaijuku backs Do Ventures to enter Vietnam’s education market: KJ Holdings, the education group’s holding company, has invested in Do Ventures Fund II, using a Ho Chi Minh City-based VC as its route into Vietnam as a shrinking population squeezes demand in Japan.

Meta names Dhruv Vohra to lead its Southeast Asia business: Vohra becomes Managing Director of Meta’s Global Business Group, overseeing commercial strategy across Indonesia, Malaysia, the Philippines, Singapore, Thailand and Vietnam as AI and chat commerce reshape online retail.

Singapore leads region with US$284B in crypto activity: Chainalysis: Singapore’s activity rose 55.4%, driven by a 94% jump in institutional trading, even as the wider Central and Southeast Asia and Oceania region contracted 6.8%. Vietnam ranked fourth with US$122.2B.

AI could lift Singapore’s potential output by 1.06%, says AMRO: Only half the gain arrives by 2035 under AMRO’s baseline scenario. Just 3.8% of firms have built AI into core operations, while 22.3% of jobs face high automation exposure, squeezing fresh graduates.

IMDA, TikTok launch programme to build Singapore microdrama talent: Professionals from 20 accredited companies will learn to produce 50-100-episode microdramas across live-action and AI-animated tracks, including ByteDance’s Seedance 2.5, before onboarding onto TikTok’s distribution ecosystem.

EV adoption spreads beyond China into Southeast Asia, says BMI: BMI expects Asia-Pacific vehicle sales to shrink 0.5% in 2026, yet EVs will keep gaining in Thailand, Indonesia, Malaysia and Vietnam. Malaysia’s tax break for imported EVs has expired, favouring locally assembled models.

AI exposure widens the credit gap across Asia-Pacific sectors: S&P: Issuers with AI demand exposure, pricing power and supply-chain flexibility look most resilient. Prolonged energy disruptions threaten manufacturing, agriculture and transport, while emerging Asia remains exposed to capital rotation.

FEATURES AND INTERVIEWS

AI agents are exposing SEA startups’ codebase problem: As developers move from AI autocomplete to agents that work on codebases unsupervised, Southeast Asia’s startups face a harder question: whether their code is in good enough shape for AI to touch safely.

Philippine brands win AI mentions but lose control of context: A Spiralytics study finds consumers increasingly ask ChatGPT, Gemini and Perplexity for buying advice, creating a new visibility problem: being named by an AI does not guarantee being described the way a brand intends.

Why token prices understate the real cost of AI agents: Founders who forecast AI costs from a model provider’s pricing page treat AI like a utility meter. Autonomous agents break that model, making usage, and the bill, far harder to predict.

Michael Padilla on turning risk management into a business: After 31 years in the US Army, the retired colonel founded Al Thuraya Holdings, now 18 companies spanning risk, security, technology and consulting. He argues the costliest failure is refusing to stop when strategy no longer holds.

INTERNATIONAL

OpenAI in talks to raise US$30B at a US$1.4T valuation: The pre-IPO round would bridge OpenAI to a delayed 2027 listing. Run-rate revenue jumped 70% since July to US$40B in August after a refocus on coding, according to Bloomberg.

SoftBank completes final US$10B OpenAI investment: The final tranche closes out SoftBank’s OpenAI commitment as the ChatGPT maker lines up a fresh pre-IPO round ahead of its planned 2027 public debut.

CYBERSECURITY

OpenAI apologises after its AI agents breached Australian govt sites: An experimental model accessed a Services Australia system holding Medicare data in June, but authorities learnt only on 10 September. OpenAI says no personal records were accessed and is setting up an independent Australian task force.

89% of Singapore leaders fear data sovereignty failures: Yet 81% of Singapore enterprises lack a formal data sovereignty strategy, the highest of eight markets in Everpure’s survey, and 68% have no plan for geopolitical data exfiltration or service disruption.

ThreatBook buys AI penetration testing platform CyberStrikeAI: The Singapore- and Hong Kong-based firm plans an Asia-Pacific rollout of an enterprise edition with on-premises deployment and audit trails this month. Researchers have linked the open-source tool to infrastructure tied to FortiGate targeting.

SEMICONDUCTOR

Infineon opens US$1.4B Thailand plant as Bangkok pushes into chips: The German chipmaker’s highly automated backend facility in Samut Prakan makes power modules for EVs, renewables and industry, anchoring Thailand’s bid to draw at least 500 billion baht in chip investment by 2029.

Taiwan’s AVC plans US$500M Vietnam expansion and R&D centre: The AI-server cooling supplier would lift its total Vietnam investment above US$1.26B with the 2027 spend in Bac Ninh, where the new R&D centre will train semiconductor and high-tech workers.

Amazon seeks to offload US$8B of Nvidia chips to investors: FT: Amazon is looking to sell US$8B worth of Nvidia chips to outside investors, the Financial Times reports, as hyperscalers hunt for new ways to finance the soaring cost of AI compute.

Malaysia needs 50,000 engineers but graduates only 5,000 a year: MBSB Research warns the talent gap will cap tech-sector earnings as Malaysia moves from back-end assembly into IC design and advanced packaging. The National Semiconductor Strategy targets 60,000 skilled engineers by 2030.

AI

OpenAI cuts ties with three safety researchers, WSJ reports: OpenAI says the trio mishandled sensitive information by sharing it with an outside AI safety group. The exits follow reports that executives brushed aside staff safety warnings, and a string of rogue-agent incidents.

Grok reportedly shaped Trump’s thinking before the Venezuela raid: Trump spent hours asking Musk’s chatbot in December 2025 how Venezuelans would react to Nicolás Maduro’s capture, Time reports. The Pentagon has since tapped Musk to co-lead a study on battlefield technology.

ChatGPT now lets shoppers virtually try on clothes: OpenAI has added a virtual try-on feature to ChatGPT, pulling the chatbot deeper into shopping as AI assistants compete to become the starting point for online retail.

THOUGHT LEADERSHIP

The cross-border due diligence questions most founders can’t answer: Chris Chen recalls a polished Singapore semiconductor pitch that faltered on three questions, starting with where the intellectual property legally sits. A clean deck and cap table cannot hide a messy cross-border structure.

What TaniHub teaches Indonesia’s agritech founders about credit: Bobby Yulandika Putra revisits the first agritech generation, from TaniHub to Crowde and iGrow, which tried to do for smallholder farmers what fintech did for SME credit, and why credit became the sector’s paradox.

Why con artists get the meeting that honest founders can’t: David Kim uses Elizabeth Holmes’ 10-minute slot with George Shultz, which ran two and a half hours, to examine why charisma opens investor doors that substance alone often cannot.

SMEs adopt AI four times slower than big business: Alexey Navolokin cites IMDA data showing SME AI adoption at 14.5% against 62.5% for larger firms, and argues that closing the gap starts with something basic: the PC on every employee’s desk.

The next information advantage is knowing what changed: Jeff Chong argues AI summarisation solves only part of the problem for decision-makers. The real edge lies in tracking what has shifted between one announcement, report or filing and the next.

Why AI could unbundle the beauty industry: With contract manufacturers and e-commerce already in place, John Millar argues AI removes one more reason new beauty brands need their own factories, R&D teams or nationwide retail networks.

Why social commerce seeding is replacing influencer retainers: With acquisition costs on Meta, TikTok and Google up 60-80% in 18 months, Serge Berezhnoy says brands are swapping guaranteed upfront fees for performance-linked creator seeding.

A 20% stablecoin APY may be less magical than it sounds: Drawing on a costly 2022 lesson, Douglas Gan urges Singaporeans used to 2.5% CPF returns to ask who pays the yield, and why, before chasing double-digit stablecoin returns.

Tokenisation: Crypto’s symbolic walk on Wall Street: Yiwei Wang examines the SEC’s Innovation Exemption for tokenised US stocks, and warns that hacks such as Bitget’s US$387.5M loss show integration with traditional finance is no substitute for security.

The sovereign shift: Why nation states trade gold for Bitcoin: Anndy Lian points to sovereign wealth fund allocations, new institutional products and ETF inflows as evidence that Bitcoin is maturing from a speculative retail asset into an institutional cornerstone.

Will Bitcoin’s 19% Uptober average survive US$100 oil?: Anndy Lian’s data points to neither a triumphant rally nor a collapse this October, but a tougher middle path as the economic pressures that defined the year weigh on crypto.

Could the US jobs report make or break Bitcoin’s rally?: Bitcoin rose 1.7% to US$85,002 and the crypto market reached US$2.89T ahead of the data, with Anndy Lian crediting institutional optimism and policy progress for the coordinated advance.

Why a good product isn’t enough for sustainable growth: Reem Bekdash argues that when growth stalls despite customer interest and a hard-working team, the problem usually lies outside the product itself.

Building a Bangladesh agency from the inside: Part 2: Tajul Islam picks up Ngital’s story after its early years, exploring what client trust looks like once work is underway in an industry he says is built on vanity metrics.

The post Ecosystem Roundup: MoneyHero’s activist wants a sale, but Richard Li holds the votes appeared first on e27.

Walk into any pitch meeting, attend any LP conference, or read any VC blog, and you’ll be handed the same well-worn gospel: build diverse teams, move fast, be transparent, hire for culture, flatten hierarchies, and always be learning. This is the catechism of the modern venture capitalist, wisdom dispensed freely, with conviction, to the founders who come hat-in-hand seeking capital.

But look at how venture capital firms actually run themselves, and something curious emerges. The gap between the advice VCs give and the organisations they operate is not a rounding error. It is structural, persistent, and, frankly, a little embarrassing for an industry that prides itself on pattern recognition.

The team building contradiction

Nothing is more sacred in VC doctrine than the team. “We invest in people, not ideas.” “Team is everything at the early stage.” Every investor says some variation of this. The data backs it up too, diverse founding teams with more than one gender or ethnicity have been shown to generate returns up to 30 per cent higher in multiples on invested capital compared to homogenous ones. Ethnically diverse founders also enjoy an average exit multiple roughly 30 per cent higher than solely white founding teams.

VCs know this. They cite it often. But here is where the cognitive dissonance begins.

Black investors make up just four per cent of venture capitalists in the United States, with only three per cent holding key decision-making roles. Black women comprise a mere one per cent of the US venture community, this in a country where Black Americans represent about 14 per cent of the population.

Only three per cent of VC partners in the US are Hispanic or Latinx, and just 14 per cent of partners are women.

And it isn’t just a US phenomenon. In the UK, 78 per cent of senior VC roles are held by men. In 2023, startups with all-women founding teams raised just three per cent of European investment dollars.

In 2024, all-female founder teams received less than 1 per cent of capital in Europe.

So while VCs tell founders that diverse teams build better companies, and the research agrees, the firms making those calls are themselves overwhelmingly homogeneous. There is a striking contradiction here: venture capital thrives on new ideas, yet many VC teams lack the very diversity they know drives innovation.

There’s even a term for it now: the “mirrortocracy.” This pattern tends to fund entrepreneurs who resemble the investors themselves, which often means white men from prestigious universities. This similarity bias restricts the variety of innovations entering the market, as ideas outside the familiar comfort zone of investors get overlooked or undervalued.

Harvard Business School professor Paul Gompers, one of the world’s leading researchers on VC dynamics, has studied this extensively. His research found that VC firms which increased the proportion of female partner hires by 10 per cent saw, on average, a 1.5 per cent increase in fund returns annually and 9.7 per cent more profitable exits. The financial incentive to diversify internally is unambiguous. And yet, little has changed.

‘Move fast’: Unless you’re the one moving

Another staple of VC advice is operational agility. Founders are told to ship fast, iterate quickly, and ruthlessly cut what isn’t working. The VC mindset, as Harvard Business Review described it, is characterised by “the individual over the group, disagreement over consensus, exceptions over dogma, and agility over bureaucracy.”

Inspiring stuff. Now consider what’s actually happening inside large VC firms.

Also Read: Growth at gunpoint: Why VCs share the blame for startup fraud

Reasons for the wave of senior partner departures from large firms include a perception that decision-making at many large firms is slow, and that funds are focused on increasing the pools of money they manage rather than working with promising founders. One investor quoted by the outlet described a cohort of top VCs who are “now deciding that they don’t want to be asset managers… they don’t want to deal with the bureaucracy or corporate structures.”

This isn’t a fringe view. What began as a trickle of departures from heavyweight VC funds about a year ago accelerated into a stream throughout 2024 and into 2025, with senior partners quitting what were long considered lifetime positions. The irony is rich: the same firms whose partners lecture founders about avoiding bureaucracy have become so bureaucratic that their own talent is fleeing.

At large VC firms, decision-making power is concentrated at the top. Junior investors often spend years sourcing deals and supporting due diligence without ever leading an investment, with real influence over which startups get funded remaining with senior partners. That is not the flat, empowered, move-fast culture they preach to founders.

The transparency problem

VCs consistently advise founders to be transparent with their boards, investors, and even the public. Radical candour, open-book management, honest retrospectives, these are standard talking points in any VC-founder conversation.

Yet the VC industry has long been defined by its opacity. VC firms are notoriously secretive. Research examining court rulings that forced some large public limited partners to disclose return information found that, in response, the most successful VC firms dropped those public LPs and replaced them with private and foreign LPs not subject to disclosure requirements.

Historically, many firms have held their cards close to their vests, and lots of VC websites remain sparse. The advice they give to founders, be open, share your numbers, build trust through transparency, is advice many are unwilling to apply to their own operations or fund performance.

California has had to step in legislatively. In 2023, California enacted Senate Bill 54 to promote transparency in venture capital funding by requiring firms with a California nexus to report demographic data on the founding teams of their portfolio companies. In other words, the state had to mandate the kind of accountability that VCs freely recommend to their portfolio companies.

‘Hire for culture’: But what culture, exactly?

Culture fit is a phrase VCs adore deploying in founder conversations. Build a strong culture early. Hire people who embody your values. Get the team chemistry right. All valid advice.

But the culture within many VC partnerships is one that outsiders rarely get to scrutinise. VC firms went on a premature hiring spree during the boom years and found themselves bloated with too many investors, in some cases resulting in underqualified investors guiding founders. The very excess and poor hiring discipline that VCs warn founders against played out at the firms themselves.

Also Read: When startups fail, should VCs go to jail?

Harvard Business School’s Gompers recommends that firms looking to build diversity must first become aware of their own biases. “Most people aren’t bad people, but we have these internal biases to think that people who look like us are smart and capable,” he says. Creating a diverse company has to start from day one. “Some firms say, ‘I’m worried about survival, I’ll worry about diversity later.’ But it’s really hard to start from a frat boy startup culture and move to one that is open and inclusive.”

That observation, made about startups, applies with equal force to the VC firms themselves.

Why does this gap persist?

Several structural forces entrench the inconsistency.

Accountability asymmetry. Founders are accountable to their VCs on a quarterly basis, burn rates, headcount, KPIs, board approvals. VC firms are accountable to their LPs on a much looser cadence, and the internal workings of a partnership are almost never scrutinised with the same rigour that founders face. Almost all VC firms do not have diverse or inclusive teams internally, so their ability to directly help their portfolio companies with these challenges is undermined.

The LP concentration problem. The top 30 funds secured 75 per cent of the year’s total VC fundraising in 2024, with just nine funds raising 46 per cent. Andreessen Horowitz alone captured roughly 10 per cent of the entire year’s capital. When capital concentrates so heavily in a small number of incumbents, the competitive pressure to reform internal practices is minimal. Established firms don’t need to change to attract LP capital.

Pattern matching rewarded, not penalised. Warm introductions to VCs are the most likely route to getting funded, and some VCs specifically state they only take meetings through warm introductions. This self-reinforcing network means the same demographics cycle through, and nobody at the top of the system has a strong incentive to interrupt it.

Also Read: The VCs writing off Indonesia are making a US$300B mistake

What should change

The good news is that this gap is increasingly acknowledged, and some firms are beginning to act.

Project Include has argued that VC firms should take the lead in tech diversity and inclusion, holding their portfolio companies accountable by adding comprehensive diversity metrics to their quarterly reporting, and implementing the same recommendations in their own firms. Taking the lead to make their teams diverse and inclusive would give them the credibility and experience to be better advisers to tech startups.

The NVCA-Deloitte Human Capital Survey has called for transparent baseline measurement across the industry. “Transparency is a powerful force for change, and we now have a clear benchmark by which we can measure progress,” the NVCA’s Bobby Franklin noted.

The market may also correct some of this naturally. Newer “next gen” VC firms that eliminate bureaucratic layers often report higher portfolio company satisfaction with responsiveness compared to traditional fund relationships. As founders become more sophisticated about the kind of investor they want on their cap table, the demand for VCs who actually practise what they preach will grow.

The bottom line

There is nothing inherently wrong with venture capitalists giving founders advice about team building, culture, speed, and transparency. Much of that advice is genuinely good. The problem is the credibility gap that opens up when the advice-giver doesn’t take their own medicine.

Founders are expected to operate under intense scrutiny and perform against explicit metrics. They face board pressure, investor expectations, and public accountability. VC firms, structurally insulated from similar scrutiny, have had the luxury of preaching without practising.

That luxury is eroding. Between LP pressure, regulatory mandates, departing senior talent, and founders who are increasingly savvy about evaluating their investors, the pressure to close the advice gap is building.

The best founders now do reference checks on their VCs. Perhaps it’s time for the industry to start doing the same on itself.

—

The post The advice gap: Why VCs don’t practice what they preach appeared first on e27.

Singapore’s early bet on artificial intelligence could give its economy a modest but meaningful lift as population ageing starts to bite. But the real test will not be how many companies say they use AI. It will be whether they can rebuild work around it.

The ASEAN+3 Macroeconomic Research Office (AMRO) said in its annual consultation report on Singapore that AI adoption could raise the city-state’s potential economic output by 1.06 per cent in the long run. That would offer a partial buffer against slower labour-force growth as Singapore’s population ages, although the transition may also create pressure for some workers, especially fresh graduates entering white-collar jobs.

The estimate is not a forecast. AMRO described it as a calibrated counterfactual focused on domestic supply-side effects. It does not include possible upside from external demand for Singapore’s AI-linked trade and investment, nor broader gains from innovation or the creation of entirely new types of work.

Also Read: iWOW’s US$11M placement tests investor appetite for Singapore’s ageing economy

Even so, the finding matters. Singapore’s long-term growth model has always relied on moving up the productivity curve, given its limited land, small domestic market and tight labour supply. AI is now being folded into that familiar national problem: how to make a small workforce produce more value.

Under AMRO’s baseline scenario, only about half of the potential-output gain from AI will be realised by 2035. The boost is expected to rise to 0.65 per cent by 2040 and 0.78 per cent by 2050. The drag from ageing is expected to become more pronounced from 2030, making productivity improvements more important to sustaining economic momentum.

Adoption is rising, but depth remains shallow

Singapore has moved earlier than many of its neighbours in treating AI as an economy-wide priority rather than a narrow technology-sector issue. The government has pushed national AI strategies, industry programmes and skills initiatives, while major cloud and chip players have continued to expand regional operations from the city-state.

Business adoption is now accelerating, but it remains uneven.

According to the Infocomm Media Development Authority, AI adoption among small and medium-sized enterprises(SMEs) more than tripled to 14.5 per cent in 2024 from 4.2 per cent in 2023. Among non-SMEs, adoption rose to 62.5 per cent from 44 per cent over the same period.

The gap becomes starker when company size is considered. The Ministry of Manpower’s (MOM) 2026 establishment survey found AI adoption at 23.9 per cent among firms with fewer than 25 employees, compared with 76.4 per cent among firms with more than 500 employees.

That divide is familiar across Southeast Asia. Larger companies are more likely to have the budgets, data infrastructure and technical teams needed to experiment with AI, while smaller firms often struggle with implementation costs, data readiness, governance concerns and a shortage of in-house expertise.

But AMRO’s report suggests the larger issue is not initial adoption. It is integration.

Only 3.8 per cent of firms surveyed by MOM had integrated AI into their core operations. Most were still planning, testing or using basic tools. Among firms already using AI, off-the-shelf generative AI tools accounted for 84 per cent of adoption, while fewer than half had implemented customised or proprietary AI solutions.

Also Read: SIA has scaled AI. Aviation must now govern the point of action

That distinction matters. A company using a chatbot to draft emails is not the same as one redesigning customer service, compliance, logistics, finance or product development around AI-enabled workflows. The first may save time at the margins. The second could change productivity.

White-collar work faces the biggest adjustment

AMRO also warned that Singapore’s labour market has a relatively high share of jobs exposed to AI compared with other high-income economies.

It estimated that occupations with a high degree of automation exposure account for 22.3 per cent of employment. These are roles where AI could substitute for a substantial share of existing tasks and potentially reduce labour demand. The exposure is concentrated in professional jobs such as financial analysis, business administration, sales, marketing and public relations.

That does not mean a fifth of jobs will disappear. Many roles are bundles of tasks, and AI may automate routine activities while leaving humans to handle judgement, client relationships, decision-making, compliance and complex problem-solving.

Business administration professionals, for instance, may use AI for data processing or preliminary analysis while continuing to interpret findings, manage stakeholders and ensure regulatory requirements are met. In this sense, AI may change the route into expertise rather than eliminate the need for it.

Firm-level evidence so far points more towards restructuring than mass displacement. Among firms adopting AI, 18.9 per cent reported job redesign and 13.9 per cent reported creating AI-related jobs. Only 6.2 per cent reported reducing headcount.

This pattern will be closely watched across Southeast Asia, where governments are trying to raise digital productivity without worsening job insecurity. Singapore may be better placed than most economies in the region because of its training system, fiscal capacity and concentration of higher-value services. But it is also more exposed, because many of its workers are in precisely the professional roles that generative AI can affect first.

Fresh graduates may feel the pressure first

One of the sharper warnings in AMRO’s report concerns younger workers.

Entry-level professional jobs often involve routine research, coordination, documentation and analysis. These tasks have historically helped graduates learn the basics before moving into higher-value work. If AI absorbs more of that junior work, employers may need fewer fresh hires, or expect them to contribute at a higher level from day one.

AI could also widen competition for some service-sector roles. By reducing language barriers and making remote delivery easier, it may allow companies to source certain tasks from outside Singapore more easily.

Also Read: Malaysia wants 300,000 AI jobs by 2030. Talent will decide if it gets there

AMRO was careful not to claim that AI is already driving graduate employment weakness. It noted that the proportion of tertiary graduates in the labour force who were employed declined in 2025, while youth unemployment stayed at 6.6 per cent in both 2024 and 2025. The share of young people not in education, employment or training improved to 3.8 per cent in 2025 from 4.1 per cent in 2024.

Graduate hiring also slowed across advanced economies during the same period, particularly in technology and professional services. Still, the report’s warning is clear: the transition costs may be concentrated among those with the least workplace experience.

From AI access to AI productivity

AMRO identified three main channels through which AI could raise potential output: higher total factor productivity as tasks are automated, additional physical capital investment as output rises, and improved effective labour input if AI helps less experienced workers learn faster.

Most of the expected long-term gain comes from productivity and capital, which together contribute 0.98 percentage points to potential output. The human-capital learning channel adds 0.08 percentage points under the baseline scenario.

The overall gain could range from 0.47 per cent to 2.10 per cent, depending on productivity assumptions. Speed also matters. If diffusion is completed by 2030, AMRO estimates a long-run gain of 1.13 per cent. Under the baseline 2035 timeline, the gain is 1.06 per cent. Under a slower 2040 timeline, it falls slightly to 1.01 per cent.

For Singapore, the policy implication is straightforward but difficult to execute. Support cannot stop at encouraging firms to “adopt AI”. It must help them apply the technology to real operational bottlenecks.

That means sector-specific use cases, implementation help for SMEs, better data systems, cybersecurity, cloud access, computing capacity and managerial capability. It also means tracking outcomes such as productivity gains, deeper integration and worker transitions, rather than counting AI users.

Also Read: Seoul tops global AI implementation as Singapore enters top tier

Existing programmes such as SkillsFuture AI courses, the SkillsFuture Workforce Development Grant and the National AI Impact Programme could help. But AMRO said training and job-redesign support should focus more closely on novice and entry-level workers in roles where AI is likely to complement human tasks.

Singapore’s AI challenge is therefore less about ambition than execution. The country has the strategy, infrastructure and policy machinery to move early. The harder question is whether enough firms, especially smaller ones, can turn experimentation into everyday productivity before demographic pressures intensify.

The post Singapore’s AI dividend will depend on what happens after the pilot phase appeared first on e27.

In a world where former military officers often trade uniforms for predictable government consultancies, Michael J. Padilla chose a different path. After more than 31 years in the US Army, rising to the rank of colonel, he founded Al Thuraya Holdings, a global group that today spans 18 companies across risk management, security, technology, consulting, energy, and business services.

What distinguishes Padilla isn’t merely the scale of what he built, but the hard-earned experience behind it: decades of leading through uncertainty, operating in high-risk environments, and helping governments and multinational corporations navigate a world that rarely behaves the way strategy documents predict.

The origins and logic of Al Thuraya

Padilla’s decision to leave the military wasn’t driven by a single dramatic moment but by a slow accumulation of observations. Over three decades, he watched standards shift, organisations grow more political, and coalition partners pursue their own competing interests.

Also Read: The 3Cs+1 framework: Navigating geopolitical fragmentation as a founder

He understood discipline, planning, and decision-making under pressure. But he also saw companies making critical decisions about security and market entry without understanding what those choices looked like on the ground. Rather than take the conventional route into another government post or a large defence contractor, he wanted to build something independent, rooted in practical experience and commercial discipline.

That logic explains why Al Thuraya sprawls across security, energy, technology, and consulting without looking like diversification for its own sake. The real connective tissue, Padilla explains, is risk. In the markets where his clients operate, a security issue quickly becomes an operational problem, which becomes a financial one, and a technology weakness can just as easily become a security vulnerability.

Humans, he notes, are often the risk themselves, through poor judgement, inaction, or simply misreading an environment. Most of the group’s businesses emerged organically, as one client problem exposed another next door.

Even something as simple as counting the group’s holdings reveals this philosophy. Public materials have cited both 17 and 18 companies, a discrepancy Padilla attributes to a document that wasn’t updated after a new entity was created. But he pushes back against vanity metrics altogether, arguing that wholly owned operating companies, joint ventures, and purpose-built entities don’t all carry equal size or strategic weight. He’d rather be precise about what’s actually functioning than inflate a number for appearances.

Also Read: The founder-to-minister pivot isn’t the problem, ASEAN’s missing governance infrastructure is

Looking ahead, Padilla is explicit that the next chapter isn’t about acquiring more companies. The priority is deepening capability where it already exists, strengthening management, and applying technology more intelligently.

Expansion will be selective, and only where genuine client need justifies it, as seen in his cautious approach to Southeast Asia, where work has followed existing clients rather than a deliberate strategic push. He’s equally willing to sell or restructure businesses if another owner could scale them faster, rejecting the founder’s instinct to cling to every asset out of sentiment.

Leadership forged between two worlds

One of the clearest illustrations of Padilla’s evolution as a leader came early in Al Thuraya’s history, when an employee was kidnapped. His instinct was purely operational: mobilise, identify local influencers, and act decisively. The client, however, saw the crisis through a different lens entirely, one bound by insurance protocols, legal counsel, and HR procedures.

Neither instinct was wrong; they were simply different frameworks for the same emergency. Padilla borrowed a principle from his military past — that the enemy always gets a vote — and applied a civilian corollary: the client gets a vote too.

That recalibration extended to how he manages information. One Special Operations habit he had to consciously abandon was compartmentalising knowledge on a need-to-know basis. In the military, restricting information protects missions; in business, it strangles ownership. Employees who don’t understand why a decision was made can’t be expected to take real responsibility for executing it.

His early international experience taught a parallel lesson: Western assumptions about authority and process don’t automatically translate abroad. Seniority on an org chart doesn’t always equal influence, and in many markets, relationships must be established before transactions can follow.

For Padilla, authentic leadership began only after he left the structured military hierarchy behind. Without rank to command compliance, he found himself personally handling sales, accounting, and project management in Al Thuraya’s early days, a grounding experience that taught him people follow credibility, not titles. Leadership, he says, is the relationship between judgement and responsibility: making a call, explaining it, owning the outcome, and reversing course when proven wrong.

That same unsentimental logic shapes his stance on governance. Padilla has not separated the chairman and CEO roles, a structure he attributes to Al Thuraya still being founder-led, but he’s open to changing it if scale or institutional investment ever demands it. “I did not create the group to protect a title,” he says.

Resilience, risk, and moral boundaries

Padilla is sceptical of companies that call themselves resilient, only to evacuate staff or freeze operations at the first sign of trouble. For him, resilience isn’t a personality trait; it’s an engineered system, built by confronting uncomfortable questions long before a crisis arrives.

What happens if a major client disappears? If banking access or communications suddenly fail? The answers, he argues, aren’t found in motivational speeches but in cash discipline, cross-trained staff, and contingency planning done while decision-makers still have time to think clearly.

Also Read: The cloud is just someone else’s computer. Sometimes that computer gets hit by a drone

The same discipline applies to Al Thuraya’s own exposure. Operating in difficult environments by design means absorbing political, economic, and security risk, but Padilla draws a firm line between advising a client and making their decisions for them. The group’s responsibility is to deliver accurate, honest analysis; the client decides how much risk to accept.

Operating across governments with varying human rights records forces constant ethical calibration. Padilla is unambiguous: Al Thuraya will not knowingly provide services aimed at civilians, unlawful repression, or objectives it cannot defend. Legality, he insists, isn’t the only test. Something can be perfectly legal and still not be work he wants the company’s name attached to. His personal benchmark is simple: could he honestly explain the engagement to his employees and to himself? If not, the answer is no.

Reflecting on where his judgement has failed him, Padilla points not to fundamental misreads of dangerous environments, but to the gap between what his teams observe on the ground and how clients choose to interpret that intelligence. He’s also candid about his own missteps, chiefly letting the attractiveness of an opportunity outpace its commercial fundamentals. Real failure, in his view, isn’t closing a company when conditions shift; it’s continuing to fund one out of stubbornness rather than strategy.

Technology with ground truth attached

Padilla’s approach to artificial intelligence mirrors his broader philosophy: useful when paired with human judgement, dangerous when substituted for it. Al Thuraya deploys AI for open-source intelligence, anomaly detection, due diligence, and forecasting, but always alongside people actually embedded in the markets being analysed (Padilla himself lives in North Africa).

He remains openly sceptical of predictive threat-scoring claims, viewing much of that language as marketing ahead of operational reality. His test for any technology is blunt: does it improve a decision, an outcome, cost, or safety? If not, it’s just another slide in a presentation.

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

That caution extends to autonomous decision-making. Padilla is comfortable with AI flagging anomalies but uneasy about allowing unreviewed systems to make calls affecting someone’s liberty, safety, or reputation. Speed, he argues, is no justification for removing accountability; a bad process made faster is still a bad process. The more powerful AI becomes, he insists, the more important human responsibility becomes.

The measure of what comes next

Asked what his younger, uniformed self would make of what he’s built, Padilla doesn’t hesitate: surprise at the sheer breadth of sectors and countries involved, and unease at how much still depends on one individual. The question that once drove him “how much can we build?” has been replaced by a more demanding one: is what we’ve built strong enough to continue without me? Building the group, he reflects, was one challenge. Making sure it matures, adapts, and continues responsibly is the next one.

The post Michael Padilla: The costliest failure is refusing to stop when strategy no longer holds appeared first on e27.

In healthcare, booking an appointment is rarely as simple as finding an empty slot on a calendar. A patient may need a specific doctor, a particular procedure, a referral, a follow-up window, or a clinic that accepts certain rules around timing and availability. Much of this still sits in phone calls, WhatsApp threads and front-desk work.

Singapore-based Zeya Health is betting that this messy layer of healthcare access is where AI can create real value. The Antler-backed startup has acquired ConnectLah, a clinic automation company that helps healthcare providers manage patient communication and booking workflows through WhatsApp.

Also Read: Graas raises US$17M, acquires Trustana to build smarter retail AI agents

The deal brings ConnectLah’s product, technology and customer relationships into Zeya’s patient access platform. Financial terms were not disclosed.

Zeya said the acquisition comes after a year in which it recorded 15-fold revenue growth. The company currently works with healthcare providers in Singapore, Malaysia, Australia, Indonesia and Vietnam, as well as a hospital partner in Cambodia. It aims to reach 1,000 healthcare providers by mid-2027.

From chat automation to AI booking infrastructure

Zeya already helps clinics and hospitals manage patient enquiries, bookings, reminders and follow-ups across messaging and voice channels. With ConnectLah folded into the company, Zeya wants to deepen its ability to automate the steps between a patient asking for an appointment and a provider confirming it.

The company describes this as “agentic booking infrastructure”. In plain terms, that means building systems that allow AI agents to do more than answer questions. They must be able to understand a patient’s request, check clinic rules, match appointment types with the right practitioner, interact with existing healthcare software, and complete a booking without breaking the provider’s workflow.

“Booking an appointment should be as simple as asking for one,” said Agastya Samat, founder and CEO of Zeya Health. “The complexity sits behind the scenes: the schedules, systems and rules that determine what can actually be booked.”

That complexity is particularly visible in Southeast Asia, where healthcare systems are highly fragmented. Large private hospital groups, specialist centres, neighbourhood clinics and independent practitioners often use different software systems, communication habits and administrative processes. Many still rely heavily on WhatsApp, calls and manual coordination, especially in markets where consumer messaging apps have become the default front door for services.

Also Read: The future of healthcare AI isn’t more data. It’s better context

This makes healthcare access a difficult problem for generic AI tools. A chatbot can collect a patient’s name and preferred time, but it cannot safely confirm an appointment unless it understands the provider’s rules and can connect to the relevant system of record. Zeya’s argument is that the real opportunity lies not in another patient-facing bot, but in the infrastructure layer that lets AI work reliably with healthcare operations.

Why ConnectLah matters

ConnectLah’s focus on WhatsApp gives Zeya a practical entry point into how many clinics already communicate with patients. The company helps automate patient communication, appointment booking, reminders and front-desk workflows inside familiar messaging channels.

For Zeya, the acquisition is less about adding a standalone product and more about absorbing workflows that have already been tested in clinics. ConnectLah customers are expected to gain access to a broader set of tools connected to existing clinic systems, while Zeya expands its provider footprint.

“ConnectLah brought AI into the channels clinics and patients already use, proving the approach in real clinics and live patient workflows,” said Jules Pereira-Gomes, founder of ConnectLah, who will join Zeya as a strategic advisor.

The transition will focus on the workflows and integrations relevant to each provider, according to the company. In practice, that means Zeya will need to account for variations in how clinics handle different appointment types, practitioner availability, reminders and follow-ups.

If executed well, the combined platform could reduce the back-and-forth that often defines healthcare booking. A patient asks a question, a staff member checks availability, clarifies details, proposes a time, waits for confirmation, then sends reminders manually. For providers facing staff shortages or rising patient volumes, shaving down this coordination burden can be more than a convenience.

A regional race for the healthcare front door

Zeya is entering a crowded but still unsettled market. Across Asia Pacific, startups and incumbents are trying to own different parts of the healthcare access stack.

At the consumer-facing end, companies such as Doctor Anywhere in Singapore, Halodoc in Indonesia, Practo in India and HealthEngine in Australia have built large patient networks around online consultations, appointment discovery or health services marketplaces. Globally, firms such as Doctolib in Europe and Zocdoc in the US have shown how appointment booking can become a major platform category.

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

Zeya’s positioning is different. Rather than building primarily as a consumer marketplace, it is targeting the provider infrastructure layer: the workflows behind bookings, reminders and patient communication. That could make it complementary to some patient-facing platforms, but it also places the company in competition with clinic management systems, patient engagement tools and AI automation vendors trying to modernise healthcare administration.

The distinction matters in Southeast Asia. Many healthcare providers are cautious about surrendering patient relationships to third-party marketplaces. A tool that helps clinics keep control of their booking rules and communication channels may face less resistance than one that tries to redirect patients into a new app.

The AI assistant angle

Zeya’s longer-term ambition is to make healthcare booking workflows accessible to authorised personal AI assistants. The idea is that a patient could ask their preferred AI assistant to find and book an appointment, while the provider’s systems still enforce the rules around what can actually be scheduled.

This is still an emerging behaviour. Most patients today are not using personal AI agents to book medical appointments. But the direction is clear: as AI assistants become more capable, they will need reliable rails into real-world services. Healthcare is one of the hardest categories because accuracy, privacy, consent and operational fit all matter.

That is why Zeya’s bet is both timely and difficult. The company is not just trying to automate clinic chat. It is trying to become part of the connective tissue between patients, providers and the AI agents that may increasingly sit between them.

Also Read: The app worked, the product didn’t: Can we install judgement into AI agents?

For now, the acquisition of ConnectLah gives Zeya more product depth, more customer relationships and a faster route into clinics already using messaging-led workflows. The bigger test will be whether it can scale across countries where healthcare software adoption, regulation and clinic behaviour vary widely.

In a region where the first point of care is often still a phone call or a WhatsApp message, the company’s opportunity is clear: make booking feel simple for patients without pretending the backend is simple for providers.

The post Zeya Health acquires ConnectLah to build AI booking rails for clinics appeared first on e27.

The question on every trader’s mind as October begins is whether Bitcoin will deliver its legendary Uptober performance or succumb to the economic pressures that have defined much of this year. The answer, based on the data, is neither a triumphant rally nor a catastrophic collapse. It is something more nuanced and arguably more challenging: a month of range-bound volatility that rewards discipline over conviction.

Bitcoin trades at approximately US$83,070, wedged tightly between crucial technical structures and economic pressure points. This price level is not random. It reflects a market that absorbed significant leverage flush late last month and now depends heavily on spot order books for direction. The immediate trend hangs in the balance, and the battle between seasonality and economic headwinds has left crypto markets heavily divided as Q4 begins.

The bull case rests on a foundation of historical precedent and technical momentum. Between 2013 and 2025, Bitcoin averaged around 19 per cent gains in October and closed the month in the green 10 out of 13 times. This track record earned Uptober its reputation as a psychologically powerful sentiment driver. The momentum setup supports this narrative. BTC has logged consecutive monthly gains heading into Q4. If the market reclaims and firmly holds above US$84,000 to US$84,433, a technical path opens toward major resistance at US$87,360 and psychological levels near US$90,000.

The altcoin rotation signal adds another layer to the bull thesis. The OTHERS/BTC chart is testing resistance, and an expected rollover in Bitcoin dominance points to early liquidity rotation into majors like Ethereum, which saw US$624.1M in weekly ETF inflows. This suggests capital is beginning to explore beyond Bitcoin, a classic precursor to broader market strength.

Also Read: The sovereign shift: Why nation states are trading gold for Bitcoin

The bear case is equally compelling and grounded in present realities rather than historical patterns. October is never a guaranteed win. Last year, geopolitical and tariff threats drove a massive US$19 billion liquidation event that wiped out the Uptober narrative, leaving the month down roughly 4 per cent. That episode reminds traders that exogenous shocks can override seasonality.

The present economic picture offers several such shocks in waiting. Brent crude holds above US$100 per barrel due to ongoing conflict around the Strait of Hormuz. Energy-driven inflation is a lingering risk that feeds directly into consumer prices. US CPI sits at 3.4 per cent YoY, keeping fixed-income yields highly competitive. The 10-year Treasury yield hovers above 5 per cent, creating an explicit hurdle for risk assets. Ahead of the pivotal October 27 to 28 FOMC meeting, the market is bracing for another potential interest rate hike. These are not abstract concerns. They are concrete headwinds that constrain the upside for Bitcoin and other risk assets.

The synthesis of these opposing forces leads to a clear conclusion. Unless institutional ETF inflows dramatically surge past US$1 billion daily, the combined pressure of expensive oil, high yields, and monetary tightening will likely confine Bitcoin to a defined trading channel. A straightforward replication of the historical 19 per cent October gain is highly challenging in this environment. Instead, expect a highly volatile start to the month with major support anchoring near US$80,811 and deeper liquidity pools resting around US$74,000 to US$75,585 if economic conditions deteriorate further.

This outlook has direct implications for how participants should operate. The split between short-term leverage trading and spot positioning for Q4 requires entirely different operational frameworks given the current economic landscape.

For leverage traders, the arena is less susceptible to cascading 10 per cent flash crashes because futures open interest has levelled out around US$53 billion. It remains highly prone to stop hunting. Major options max-pain levels sit below the current price. If Bitcoin attempts to rally but repeatedly fails to break the US$85,000 resistance barrier, scaling into short positions targeting an inefficiency sweep back toward US$80,875 becomes a viable strategy. Do not chase longs inside the current cluster. Wait for a definitive daily close above US$85,000. Reclaiming this level triggers a short-squeeze vector toward US$87,397, with a final target near the US$90,000 psychological barrier.

Also Read: Can Bitcoin hold US$82,000? Inside the security fear and macro storm

For those positioning for the entirety of Q4, the entry strategy should anticipate economic friction in late October. With the 10-year Treasury yielding 5.17 per cent and oil above US$100, the market will likely experience a mid-month liquidity drain. Treat any geopolitical or economically driven pullbacks into the US$74,000 to US$75,585 demand zone as a high-probability spot-buy tier.

On the altcoin front, Bitcoin dominance remains elevated at 58.67 per cent. Capital is not yet flowing freely into high-beta assets. Keep spot allocations concentrated heavily in large-cap majors like Ethereum or Solana until Bitcoin dominance drops cleanly below 58 per cent, which will act as the green light for broader altcoin exposure.

My perspective is that the Uptober narrative, while emotionally satisfying, distracts from the structural reality. The market is not in a phase where historical averages dictate outcomes. It is in a phase where economic conditions set the boundaries, and technical levels define the trading range.

The most successful participants this month will be those who respect the range, manage risk around the FOMC meeting, and position for Q4 through patience rather than fear of missing out. The battle between Uptober and Downtober will not produce a winner in the traditional sense. It will produce a grinding, volatile month that rewards those who understand the difference between a seasonal pattern and a structural trend.

—

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 Uptober or downtober: Will Bitcoin’s 19% seasonal average survive US$100 oil? appeared first on e27.

Singapore’s position as one of Asia’s most advanced digital economies is built on a simple promise: global companies can move data, capital and operations through the city-state with confidence. A new study suggests that promise is becoming harder to keep.

Research released by data storage and management company Everpure found that 89 per cent of Singapore-based enterprise leaders believe a data sovereignty failure could cost them their jobs. The fear is not only personal. The same proportion said such a failure could damage their organisation financially and reputationally.

Also Read: The new border: Why server farms are the battleground of AI sovereignty

Yet the more striking finding is the gap between concern and action. According to Everpure’s Global Data Sovereignty Report 2026, 81 per cent of Singaporean enterprises surveyed do not have a formal data sovereignty strategy in place, the highest share among the eight markets covered in the study.

That matters because data sovereignty is no longer just a legal question about where information is stored. It has become a business continuity, geopolitical and vendor risk issue.

At its simplest, data sovereignty refers to the idea that data is subject to the laws and controls of the country or jurisdiction in which it is stored, processed or accessed. In practice, the challenge is more complicated: companies must know who can access their data, which foreign laws may apply, and whether a cloud or software provider could be forced to hand over information or suspend services during a political dispute.

For Singapore, a regional headquarters for banks, tech companies, logistics players and digital platforms, the issue cuts especially close. The country’s economy depends on trusted cross-border flows of information. At the same time, its companies often rely on global cloud, software-as-a-service (SaaS) and cybersecurity vendors whose infrastructure may span several jurisdictions.

Awareness is high, preparedness is not

Everpure commissioned research firm Vanson Bourne to survey 2,100 C-suite and IT leaders from large enterprises across the UK, France, Germany, Australia, Japan, South Korea, Singapore and India in June 2026. Singapore accounted for 100 respondents.

The study found that 93 per cent of Singapore organisations recognise data sovereignty as a business concern, compared with 90 per cent globally. Some are already changing procurement behaviour. About 41 per cent of Singapore respondents said they are limiting their use of SaaS providers that rely on non-domestic infrastructure, while 86 per cent said they would compromise on advanced features to work with a local or sovereign provider.

Also Read: Should cybersecurity be nationalised?

But recognition has not translated into operational readiness. In Singapore, 63 per cent of enterprises said they lack full visibility into who can access, control and manage their data. More worrying, 68 per cent said they have no mitigation plans for geopolitical data exfiltration or service disruption.

This is the heart of the “sovereignty gap” highlighted in the report: executives know the risk is material, but many organisations have not built the governance, technical controls or response plans needed to manage it.

Nathan Hall, Vice President and General Manager for Asia Pacific and Japan at Everpure, said the risk for Singapore lies in the disconnect between digital maturity and organisational preparedness.

“Singapore is one of the most digitally mature markets in the world, yet 81 per cent of enterprises here are operating without a formal data sovereignty strategy. That gap between awareness and action is the real risk,” he said. “Sovereignty is not simply about where data sits — it is about knowing who can access it, which jurisdictions apply, and whether critical services could be disrupted.”

The point is especially relevant in Southeast Asia, where regulation is still uneven across markets. Singapore has a mature data protection regime under the Personal Data Protection Act, while neighbouring economies are developing or refining their own privacy, cybersecurity and localisation rules. For regional companies, this creates a patchwork problem: data may be generated in Indonesia, processed in Singapore, analysed through a US-headquartered SaaS platform, and stored on infrastructure distributed across several markets.

The AI factor

The sovereignty question is becoming more urgent because of artificial intelligence. As companies feed more enterprise data into AI systems, the boundaries around storage, access and reuse become harder to track. Sensitive operational data may move into model-training environments, analytics platforms or third-party applications without executives fully understanding where it goes or how it is governed.

Also Read: AI governance is moving from promises to proof

This is not just a theoretical risk. Banks, insurers, healthcare groups and government-linked enterprises in Southeast Asia are under growing pressure to adopt AI while maintaining strict controls over customer data. For startups and scaleups, the challenge is different but no less serious. Many depend on global cloud platforms and AI tools from day one, often without the resources to conduct deep vendor risk reviews.

Everpure’s survey suggests that companies are still treating sovereignty as an extension of cybersecurity or compliance. That may be too narrow. Cybersecurity focuses on preventing unauthorised access or attacks. Compliance focuses on meeting legal obligations. Sovereignty adds another layer: whether an organisation retains effective control over its data when foreign laws, vendor dependencies or geopolitical shocks come into play.

Rahiel Nasir, Research Director and Lead Analyst for Worldwide Digital Sovereignty at IDC, described the shift as a board-level issue. “The challenge for executives is not just about knowing where their data are hosted,” he said. “It is about being in total control of all data access and transfers, including all metadata, guaranteed protection against extra-territorial data requests, and managing IT and vendor risks in the light of geopolitical uncertainties.”

From compliance checklist to operating model

Everpure argues that companies should move towards “sovereignty by design”, where governance and controls are applied based on the risk of each data set, application and workload. In practical terms, that means mapping critical data, classifying it properly, understanding vendor access, and deciding which workloads require stricter controls.

The distinction matters. Not every piece of enterprise data needs the same level of protection. A marketing dashboard, payroll file and national infrastructure system carry different risks. A blanket localisation strategy can be expensive and restrictive; a purely global cloud approach can leave companies exposed. The harder but more useful path is to decide which data must remain under tighter corporate control and which can safely sit within global platforms.

Also Read: Southeast Asia’s AI buildout is racing toward a power wall

For Singapore, the findings should be read less as an indictment and more as an early warning. The country has spent years building itself into a trusted digital hub for Asia. Maintaining that position will require not only strong national regulation, but also stronger internal discipline among enterprises using cloud, SaaS and AI systems.

The boardroom fear captured in Everpure’s report may sound dramatic. But in a region where data flows underpin finance, trade, healthcare and digital services, sovereignty failures are no longer abstract policy debates. They are operational risks, and increasingly, leadership risks.

The post Why Singapore firms fear data sovereignty failures but remain underprepared appeared first on e27.

Indonesia’s agritech sector looks different in 2026 than it did when TaniHub raised its Series B in 2021. The cohort of platforms that emerged in the late 2010s — TaniHub, Crowde, iGrow, Sayurbox, and others — aimed to do for smallholder agriculture what fintech had done for SME credit: build technology infrastructure for a segment the formal banking system could not serve at scale.

Some platforms have grown. Several have restructured significantly. A few have wound down. The credit arm of the sector, in particular, has been through a harder cycle than founders or investors anticipated.

After fifteen years inside Indonesian risk functions, I have watched this cycle with direct visibility into the structural risk questions it raised. The story underneath the platform-level outcomes is more useful than the company-by-company narrative.

What the first agritech generation built

The Indonesian agritech sector that emerged after 2016 built two adjacent infrastructures. On the commerce side, platforms built marketplaces connecting smallholder farmers directly to institutional and retail buyers — restaurants, hotels, supermarkets, modern trade. On the credit side, P2P lending platforms targeted agricultural working capital specifically — seeds, fertiliser, equipment, harvest financing.

TaniHub, founded by Pamitra Wineka, Ivan Arie Sustiawan, William Setiawan, and Michael Jovan in 2016, was one of the most prominent platforms to combine both sides through its TaniFund credit arm. Its Series B in 2021 was one of the largest agritech rounds Indonesia had seen.

Also Read: Why Southeast Asian agritech must build for acquisitions, not IPOs

What turned out to be harder than expected

Three things proved structurally more difficult than early models priced in.

• Harvest cycle credit timing. Agricultural loans do not behave like consumer or SME credit. They are paid back from a single harvest event that can be three, six, or twelve months out, and that may or may not arrive at the expected volume or price. A platform’s cash flow timing assumes regular monthly repayments. A farmer’s cash flow timing does not.
• Default correlation. Agri-credit defaults are not independent in the way SME or consumer defaults mostly are. A drought, a flood, a pest event, a commodity price collapse — any of these correlate defaults across hundreds of borrowers in the same region simultaneously. Models that treat each loan as independently distributed misprice this correlation, sometimes by an order of magnitude.
• Operational cost per loan. The cost of underwriting, monitoring, and collecting a small agricultural loan — often spread across remote geographies, often involving in-person verification — is high relative to the loan ticket. Platforms that priced loans against urban operational assumptions ended up subsidising rural credit from other revenue lines.

Lessons learned

Six principles emerge from the Indonesian agritech credit cycle.

• Agri-credit is not consumer credit with mud. The underlying risk physics — harvest cycles, weather correlation, commodity volatility — is structurally different. Models built for one will misprice the other.
• Correlation is the silent killer. The single biggest pricing error in early agritech credit was underestimating how correlated defaults can become inside a single weather or price event. Diversification across crops, geographies, and harvest cycles is solvency infrastructure, not a marketing point.
• Operational cost is destiny. Platforms that did not build for the cost of remote, small-ticket lending from day one ended up cross-subsidising it forever — or stopped lending. There is no version of agritech credit at scale without operational cost discipline designed for the segment.
• Funding tenor must match harvest tenor. Short-term retail or institutional funding does not pair well with agricultural cash flows. The platforms that survived better had funding partners willing to hold positions across full agricultural cycles, not calendar quarters.
• The buyer side is more durable than the credit side. The commerce infrastructure built by Indonesian agritech — connecting farmers to buyers, aggregating produce, building cold chain — has aged better than the credit infrastructure built alongside it. Future capital allocation should reflect that.
• The credit gap is durable, the model needs to mature. Indonesian smallholder agriculture will still need credit access whatever the platform-level outcomes of the first generation. The next models will need to absorb the lessons the first generation paid for, or they will pay for them again.

Also Read: Agritech investors are learning that infrastructure matters

The macro stakes

TaniHub and the agritech cohort built something Indonesia had not had before: a market-facing technology layer on top of smallholder agriculture. Not all of it survived. The parts that did, and the lessons from the parts that did not, are now the foundation for whatever comes next.

Indonesian smallholder agriculture employs tens of millions of people and produces a significant share of the country’s food supply. The credit gap at the farm gate is one of the most important development finance questions in Southeast Asia. The first generation showed that the gap could be addressed by technology platforms — and that it is harder than the original models predicted. The next generation has the data to do better. The opportunity is still there.

—

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 agritech credit paradox: Lessons from TaniHub and Indonesia’s first agritech generation appeared first on e27.