Singapore just released its report on venture funding for 2025, and almost every write-up reads the same way. Funding winter. Capital’s gone quiet. Hold the line, it’ll come back.

I think that’s the wrong story.

I’ve been sitting with these numbers for a few days, and the more I look at them, the more I’m convinced we’ve been telling ourselves the comfortable version. The comfortable version is that the money left and the money will return. The harder version, the one I actually believe, is that the region made a strategy bet a decade ago, the bet didn’t have an exit attached to it, and 2025 is just the year the math stopped hiding. We’ve had a few of these years where the math stops hiding. This is another one.

So let me do a bit more opining than usual. This one’s a little spicy.

* * *

The number everyone read

The headline is genuinely rough. In 2025, Singapore recorded 472 venture deals, down 35 percent from the year before. Total capital raised came in at 4.6 billion US dollars, down 34 percent year on year. And Singapore is the strong one. Across the ASEAN-6, both deal value and deal volume hit a four-year low.

Now hold that next to the United States in the same year. Silicon Valley deal value nearly doubled, to around 160 billion dollars. A lot of that was two rounds: OpenAI at 40 billion, Anthropic at 15 billion.

Two companies, in one country, raised more than ten times what the entire island of Singapore raised across 472 deals all year.

The easy conclusion is that capital is concentrating into American AI and starving everyone else. That’s true as far as it goes. There’s real gravity pulling allocators toward the bleeding edge, and that gravity sits in Silicon Valley.

But that’s a description of the weather. It doesn’t tell you why our house is the one with the leak.

For that, you have to go back further than last year, and look at what we actually spent the money on, and what we expected to get out the other side.

* * *

The bet we made

Here’s the part that doesn’t get said enough. For most of the last decade, Southeast Asia poured its venture money into consumer. Ride-hailing, e-commerce, food delivery, the super-app. The big, beautiful, blitzscaled consumer story where you capture a young, mobile-first population of 700 million and become the thing they open twenty times a day.

I’m not mocking it. I lived through the optimism. Grab, GoTo, Sea, Lazada, Shopee. These companies built the rails the whole region runs on now. Digital payments are everywhere because of them. That’s real, and it was needed. Consumer is the precedent layer. Most maturing markets start there, build the rails, then transition. That part is natural.

But look at the allocation. In 2023, more than a third of Southeast Asian venture deal value went into consumer. The honest caveat is that “consumer” is a fuzzy line, depending on whether you fold in consumer fintech, so treat the exact figure loosely. Even on the conservative read, you land somewhere north of thirty percent. Run the same count in the US that year and you’re in single digits. The number I keep landing on is around three and a half percent.

Read that again. We put an order of magnitude more of our capital into consumer than the most mature venture market on earth did.

And we weren’t growing out of it. We were accelerating into it. Consumer’s share of regional deal value kept climbing while software’s share fell. So while the US was doing the boring, durable thing, funding enterprise software and infrastructure, we were doubling down on the consumer copycat play right as the cheap money drained out.

Why does that matter? Because of what happens at the end.

* * *

The door that was never there

Every venture dollar is a bet on an exit. Money goes in, and somewhere down the line it has to come out bigger, through a sale or a listing. No exit, no returns. No returns, no next fund.

So how did the region do on exits? Here’s the number that should be tattooed on every term sheet. Since 2015, the entire Southeast Asian venture market generated roughly 70 billion dollars in exit value. Sounds fine until you look underneath. More than 55 billion of that came from three exits, all in 2021. Stretch it out and nearly 87 percent of all exit value since 2015 came from six companies. Take it to the top twenty and you’re at 96 percent.

Yes, there’s always a power law. Concentration is normal. But strip out a handful of unicorns and the regional market has returned almost nothing to almost everyone. The investment-to-exit ratio has run consistently above twenty to one. Twenty dollars in for every dollar that found its way out.

It’s been a trap. The Hotel California of venture. You can check in, but you can never leave.

And here’s the part that connects the dots. The few giant exits we did get didn’t happen here. Grab went out via a SPAC on the Nasdaq. Sea listed on the New York Stock Exchange. They had to leave to get out. The Singapore Exchange, the biggest in the region, ranks only ninth by market value in Asia-Pacific, and several regional exchanges still carry listing rules strict enough to keep a cash-burning consumer company out entirely. For a blitzscaled consumer business, the local IPO was a closed door.

So put it together. We funded consumer companies built on the growth-at-all-costs playbook, and that playbook only pays off through a big public listing. We never built the public markets to list them on. We built companies for a door that, at home, was never there.

That’s not a winter. Winter ends. This was a design flaw.

* * *

Consumer is the hardest thing to sell, everywhere

This is the part I want founders and investors to chew on, because it goes beyond us. Consumer is one of the hardest categories to exit anywhere in the world.

Think about who actually buys companies. In enterprise software there’s a deep, permanent bench of buyers who do this all day. 2025 was the most active year on record for software M&A, with strategic buyers alone accounting for around 42 percent of deals. The most active software acquirers in 2024 included IBM, Cisco, Autodesk, Nvidia. There were 22 firms that each made at least five acquisitions in a single year. That’s a machine. A standing market of people whose job is to buy companies. What are they buying for? Recurring revenue, mission-critical, sticky, hard to rip out.

Now ask who the standing buyer is for a regional food-delivery app, or who’s lining up to roll up consumer brands in a market where customers switch the second someone else runs a discount. There isn’t a bench. Consumer internet leans almost entirely on the IPO. And we just covered what happened to that door.

Let me be fair, because the honest version is more interesting than the cheap one. Enterprise exits aren’t easy either. Only about ten percent of companies tagged as software ever get acquired. IPOs are about six percent of software exits. The median software acquisition went for roughly three times revenue, not the eye-watering multiple people imagine. B2B is not a golden ticket.

What enterprise has is a functioning market of repeat buyers. Consumer mostly has the IPO. It’s a difference in optionality, in how many doors are actually open. We bet the region on the category with the thinnest exit options, and didn’t build the one exit that category depends on until recently. If you wanted to design a liquidity crunch on purpose, that’s how you’d do it.

* * *

The people who built it are now saying it

What makes this report worth reading past the headline is the back half, where they ran candid pieces from a row of the region’s investors. To their credit, the honesty is right there.

Vishal Harnal at 500 Global names liquidity as the clearest challenge facing the region, pointing straight at underdeveloped exit markets and the long holding periods that wear founders and investors down. Angela Toy at Golden Gate is just as direct, conceding the region still lacks depth in both M&A and secondaries to get people their money out.

The one that stuck with me is from Cyril at SOSV, who lays out the question every Singapore founder eventually asks out loud. If the place you ultimately have to go for capital, scale, and an exit is San Francisco, why not just start there on day one? Why build here at all? That’s a tough one to sit with. It’s not a critic on the sidelines. It’s a GP at an active global fund saying the quiet part into a government report.

Then there’s Antler. They’ve raised about 1.5 billion dollars globally, from dozens of institutions and sovereign funds. The amount that came from Singapore institutions was around 10 million. The US allocates roughly five percent of its capital to venture as an asset class. Singapore sits well below one. So even the domestic money, the money that’s right here, mostly doesn’t back the local market. The capital sits in the city. It just doesn’t believe in the thing the city keeps saying it wants to be.

When this many people who built the market all point at the same missing piece, it stops being a complaint and starts being a diagnosis.

* * *

So what do we actually do

To be clear, Singapore isn’t sitting still. The response is substantial: an extra billion dollars into Startup SG Equity for growth-stage companies, a new 1.5 billion dollar anchor fund aimed squarely at strengthening exits, and a Singapore Exchange and Nasdaq partnership we’ve talked about here before. Almost all of it is about building the exit door now, after a decade-plus of funding companies that needed it and didn’t have it.

I’m not saying that to dunk on the policy. The policy is correct. Real liquidity, a working M&A culture, a credible place to list, that is exactly the right thing to spend on. My point is that we’re building the staircase after everyone already jumped. The companies that needed this in 2018, 2021, 2023 are gone or got out somewhere else. The question is whether the next decade of founders builds for the door that’s finally going up.

So here’s where I land. Stop building for the exit that doesn’t exist, and start building for the one that does.

That’s been our thesis at Indelible Ventures: back the higher-probability path from where the region actually is, and keep tracking how that liquidity path shifts over time. If the dependable way out is acquisition rather than a hometown IPO, then build the kind of company that has buyers. Real revenue, defensible product, something a strategic acquirer or a private equity firm actually needs to own. Not a big user number you’re hoping a public market rewards someday. Reality over vanity metrics. Capital efficiency stops being a constraint you tolerate and becomes the strategy. The companies getting funded here, and more importantly the ones that can get out, are the ones with clean unit economics, not the steepest growth chart.

I want to say something specific about the Philippines, because I’m genuinely optimistic about it and the lesson lands well there. The consumption story is real. Household spending is something like three-quarters of GDP. The young population, the digital adoption, all of it is genuine. The trap would be to look at that and run the same blitzscaled copycat playbook that just left the rest of the region holding companies it can’t sell. The opportunity is to build for that consumption with discipline, with models that travel across similar markets, and with an exit in mind from the start.

Same demand, smarter strategy. The fundamentals are a gift. The old playbook was the problem.

* * *

What it actually says

Southeast Asia’s problem in 2025 was never that it ran out of money. The region is full of money. Family offices, sovereign funds, the whole lot. The problem is that we built a generation of companies with no clean way to turn into returns, in the category least likely to produce them, listing on markets that mostly weren’t here.

That’s fixable. But only if we’re honest that it was a choice, not the weather.

The money will come back. The question is whether we’ll have built something it can actually leave through.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

I hopped into a taxi in Bangkok last week and the driver, a man north of fifty, spent the ride telling me what he was building with AI.

Not complaining about the economy. Not asking where I was from. Telling me about his project.

I’ve been turning that over ever since, because it isn’t an isolated thing. For weeks now I’ve been scanning event listings in whatever city I land in, and the pattern is hard to miss. It isn’t pitch nights anymore. It isn’t another fireside with a fund manager. It’s vibe coding meetups, agentic AI sessions, AI trainings. Paid attendance, no walk-ins, speakers who’ve shipped real apps. KL has them. Singapore has them. Bangkok and Manila have them. Go on Lu.ma or Eventbrite right now and there’s probably one happening in your city this week, maybe two.

I know this firsthand because I run some of them. I host AI salon events in Bangkok, and I’ve watched the rooms change.

So while the rest of the startup world argues about whether funding is back, looking at numbers that are frankly pretty dismal, there’s this whole other thing happening in cafes and malls across Southeast Asia. Regular people are learning to build software by talking to a machine.

Why I trust this one

I dismiss most AI hype on reflex. My feed is littered with slop, articles that read like they were generated by the thing they’re describing, people calling everything the future. I scroll past it.

This is different, and the reason is simple. People are paying to show up.

And it’s a different crowd than I’m used to seeing at startup events. University students and fresh grads who can see the job market tightening and are choosing to get ahead of the curve instead of waiting it out. Founders who can’t afford a dev team. Marketers. People with an idea and no technical co-founder, who a year ago would have been stuck with that idea trapped in their head, never seeing daylight. This is the no-code, low-code movement, upgraded and supercharged into the current AI era.

The category has a name now: “vibe coding”. I’m not a fan of the term, all that talk of vibes and feel grates on me, but it’s the vernacular, so I’ll use it. You describe what you want in plain language and the AI writes the code. That’s the whole thing.

I do it myself. I’ve used AI coding to replace most of our software stack. Thinking back to the friction of a couple of years ago versus how good this is now, and then projecting forward to how good it’ll be as the models keep improving, is genuinely one of the more interesting arcs I’ve lived through as an operator.

From apps to agents, which is where it gets serious

Building an app is one thing. The next rung up the ladder is building an agent, and agents are a different animal.

Most people, once you get out of the tech bubble, still picture a chatbot. You type, it types back. You ask, it answers. A better Google. That’s generation. It makes text, images, words.

An agent acts. It doesn’t tell you how to clear your inbox, it clears your inbox. It books the meeting. It sends the email. It runs commands on your machine. It talks to other software and gets things done with barely any input from you.

That’s the entire ballgame for risk. A chatbot needs a human to type every prompt. Every harm one causes still started with a person asking for it. An agent can plan, decide, and act on its own initiative. It can cause harm nobody asked for.

I want to be clear that I’m bullish on this. Hugely. But being bullish and being measured aren’t opposites, and the risk side of this deserves honest airtime.

Two examples everyone in the open-source world is talking about. The first is the lobster: OpenClaw. It went viral the moment it dropped. It connects an AI model to your messaging apps and acts on your behalf, books things, browses, runs commands, manages your house. People pulled their old Mac minis out of drawers to run it. Apple caught the wave and nudged the price up. It is not a Southeast Asian product, and we should be honest about that. It went viral hardest in China, which has been well ahead on the open-source movement. Southeast Asia needs to kick into gear as a fast follower, even when we’re not the origin.

The second is Hermes, out of a US research lab a few months back. What makes it different is memory. It lives on your own server, runs all the time, and gets better the longer you use it. It remembers what you told it last Tuesday. It writes down how it solved a problem so it never starts from scratch again. By this month it was the most-used agent out there by some measures, hundreds of billions of requests a day, hundreds of thousands of developers piling in within three months.

Here’s the part that should make you pause. Three separate security audits this year found malicious code hiding in the add-on skills people share for these agents. Think about what that means. An autonomous thing, running constantly, on your own machine, with access to your messages and files and maybe your ability to spend money, pulling new abilities from a community marketplace that’s already been found to contain things designed to hurt you. That isn’t a future problem. It’s a this-year problem, and it’s happening on hardware people own, in their homes, outside any IT department or compliance check.

A friend who’s far sharper than me on this put it well. Permissioning an agent is like onboarding a new intern. You give them enough access to act, but not enough to break things. If humans are entities of action, we have to treat agents as entities of action too, with the same scoping and the same limits. The catch is that getting that right still takes real technical skill, and most of the people downloading the lobster don’t have it.

So who’s writing the rules

Surely someone’s regulating this. Here’s where it actually stands, and the answer is more interesting than “nobody is.”

Three big global players, three different postures. The US is actively deregulating to keep its lead, tearing up the old safety rules and trying to stop its own states from making their own. The posture is get out of the way, though there was an executive order floated recently that would have made new models notify the government before public release, something closer to how the FDA approves a drug. It got paused, not signed. We’ll see. Europe, true to reputation, has the most serious regime, and just this month agreed to delay the hardest parts, the high-risk rules, by over a year. Competitiveness pressure. So even the strictest regulator in the world is loosening its grip right as agents arrive. And China is the strictest in practice and the only one already acting on agents specifically, real enforcement, thousands of non-compliant services shut down. Telling, the country where everyone installed the lobster also told its own government agencies and state banks not to put it on work devices. The adoption champion got nervous about its own craze.

Even the deregulating US quietly started building standards for autonomous agents. So nobody actually thinks this is fine. Everyone sees the gap. They’re just moving at wildly different speeds.

Southeast Asia is that same story compressed into one region, running at three speeds. Vietnam, maybe not who you’d guess, has the only real binding AI law here, passed late last year, enforced since Q1, risk-based with actual prohibited uses. It tracks, given how much of the region’s developer talent sits there. Singapore did something very Singapore: the world’s first governance framework built specifically for agentic AI, detailed and thoughtful, and deliberately voluntary. No teeth. The bet is give industry sophisticated guidance, remind everyone they’re still liable when their agent screws up, and keep the innovation onshore. They’ve already refreshed it with case studies from the likes of OCBC, Tencent and Workday. A living document, which is the right call given the pace. And then Malaysia, where I’m based, sitting on one of the most aggressive agent rollouts in the region, with its actual rules still in draft. Not here yet.

Here’s the whole thing in one line. Everyone, globally and right here at home, is regulating the last war. The last war was chatbots generating bad content, the stuff you can ban after it spreads. We saw it when Indonesia, Malaysia and the Philippines banned Grok over deepfakes, including images of children. Three countries, fast, coordinated, and fully deserved. But that’s the model: react after the harm, fold quickly. And every one of those images still needed a human to type the prompt.

The next war is agents taking bad actions on their own, because the black box decided that was the thing to do. That war is already shipping. Through anonymous downloads, onto personal machines, learned at meetups across the region, in a place where exactly one country has even a voluntary framework and the country with the biggest rollout is still drafting.

We’re banning the thing that needs a human to ask. We haven’t started on the thing that doesn’t.

What I keep coming back to

I’ll be honest, I don’t have a clean answer. Part of why I raised this is that it was a quiet news week. But the bigger part is that I can’t stop noticing the trend, and I doubt I’m alone. If you’re a CISO or a CTO or sitting in a compliance function, you’re already living this, because the whole enterprise is integrating more automation and more agents by the month, and the risk side is going to drag a regulatory environment into the room whether we invite it or not. It always does, the moment a technology touches enough of society. So it’s worth thinking now about what that reaction is likely to look like, instead of being surprised by it.

But I keep coming back to those meetups. To the rooms full of people building. Because that’s the real story, and it isn’t happening in a lab or a boardroom. It’s happening on your street, in cafes, in small event rooms. People in this region are adopting this faster than the people meant to govern it can keep up, and honestly that’s fine, because that’s how technology has always moved. I’m genuinely excited to see Southeast Asia stop being just a fast follower and start leapfrogging, with the macro trends, shifting supply chains, and regional growth all pointing the same way. There’s an enormous opportunity here, and I think this is going to sit at the front of it.

So I’ll leave you with the question I can’t answer myself. If you’re building with these tools right now, who’s responsible when your agent does something you didn’t intend? You? The person who built the tool? The government that hasn’t written the rule yet?

Right now the honest answer is nobody knows. And everybody’s building anyway.

As we should. But take a beat on that one.

Tell me where I’m wrong.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

Two stories from Southeast Asia this week, and almost nobody connected them.

The Philippines unveiled the marker for the Pax Silica industrial hub in New Clark City. Twenty plus companies expressed interest. A dozen are billion-dollar US firms. And on the same day, Manila publicly rejected the US request for diplomatic immunity and US legal jurisdiction over the zone. The hub will operate under Philippine law.

Malaysia rewrote the rules of how startups go public on Bursa. VC firms can act as listing agents. Retail investors can participate for the first time. A real funding escalator from regulated crowdfunding to LEAP Market to ACE Market.

But the Malaysia story is part of something bigger. Singapore signed an SGX-Nasdaq dual listing bridge last November. The ASEAN-6 signed a cross-border depository receipts MOU in December 2024. Indonesia is tightening listing rules to chase quality. The whole region is rebuilding its public markets for venture-backed companies at the same time, and almost no one is tracking it as one story.

This episode walks through the two races happening in Southeast Asia right now. The industrial race for the AI economy, and the capital markets race for venture-backed exits. Each country is making different bets. Each country is solving for a different segment. Where you build matters now in a way it didn't five years ago.

I'm bullish on the Philippines. But the country has a gap on the capital markets side, and closing that gap is the work of the next two years.

Real. Raw. Relatable.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

There’s a version of this week that looks like a good week for Southeast Asia’s startup ecosystem.

The Q1 2026 funding report shows the highest quarterly capital raised since late 2022. Malaysia’s hottest IPO in sixteen years prices and lists next week. OpenAI and Anthropic both announce major new enterprise offerings backed by some of the biggest names in global private capital.

Here’s the version where you actually read the numbers.

One data centre deal accounts for over 70% of the quarterly funding total. The chip company getting 95 times oversubscribed has three-quarters of its revenue coming from China and a tax exemption that expired eight months ago and hasn’t been renewed. And the AI labs building $4 billion services arms are, if you read what they’re actually saying, admitting that their models are not easy to deploy in the real world.

Three stories. Let’s take them properly.

* * *

The Real Q1 2026 Funding Number

DealStreetAsia dropped their Q1 2026 Southeast Asia funding report this week. It’s making the rounds. The headline: $2.81 billion raised, the highest quarterly total since Q4 2022.

One deal, DayOne, a Singapore-based data centre operator, raised $2 billion in a Series C. I’ll put a mild caveat on that: this is a data centre, not technically a startup, and it was spun off from an existing entity. It’s in the numbers because it carries a Series C label. That’s fine. But it’s worth knowing what you’re looking at.

Strip it out. You have just under 100 deals and under $800 million combined. The lowest quarterly deal count in at least eight years.

That’s the actual funding market founders in this region are navigating right now. Not the headline. The actual market.

On the Singapore Number

The report shows Singapore capturing 91.5% of total capital. I’m honestly always a little skeptical of that figure in isolation, and here’s why.

Singapore is the home of the holdco. If you’re a founder in Malaysia or Indonesia or Vietnam trying to raise international capital, you’re not going to stay registered in your home jurisdiction. You’re going to put a holding company in Singapore, because the legal and regulatory environment is cleaner, because international investors are more comfortable with it, because that’s just how it’s done. Your operating company may still be fully onshore in your home market.

So some portion of what gets reported as “Singapore funding” is actually capital going into companies operating across the region, just routed through a Singapore holdco. How much? Hard to know. But it’s worth holding that nuance when you see the 91.5% figure.

What it definitely does tell you is that the Singapore jurisdiction matters, for capital access, for legal infrastructure, for institutional credibility. That part is real regardless of the holdco effect.

Malaysia: Signal or Noise?

The report calls out Malaysia as a bright spot, ranking second in Southeast Asia by deal volume for the first time. Eighteen deals, the highest quarterly count since Q3 2024.

I’m active in the Malaysian ecosystem. My honest read: take this with some salt. When you dig into what drove the number, a meaningful portion came from small cheques through a single accelerator programme. That’s not nothing, but it’s not the same as organic deal activity across the ecosystem.

I don’t want to be the one pouring cold water on every green shoot, and I’m not saying the Malaysian ecosystem isn’t moving. But there’s a difference between an ecosystem inflection and a batch of accelerator cheques inflating a quarterly number. We’ll know more by Q3.

Where the Money Is Actually Going

If you’re a founder asking where capital is flowing: AI. Specifically agentic AI, automation of workflows, tasks that execute with limited human oversight. Not chatbots. Actual agents doing actual work.

AI and ML deals came in second by volume in Q1 with thirteen transactions. The biggest was Amity’s $100 million Series D. Worth noting: Amity has a long-standing relationship with CP Group, one of Thailand’s largest conglomerates, which is the lead investor. That context matters for how you read the round. It doesn’t diminish the achievement, it’s still a strong signal of appetite in the space, but it’s worth knowing.

The message for founders: if you’re building real enterprise automation, real measurable productivity gains, there is capital. Not a lot. But it exists and it’s consistent.

The Quiet Problem Nobody Names

There’s something that doesn’t get said clearly in this ecosystem, so let me say it.

There is a growing number of zombie companies across Southeast Asia. Not failed companies, companies that can’t raise new capital, can’t grow meaningfully, but won’t die. They exist in a kind of operational limbo. Technically alive. Burning slowly.

Part of what sustains this is that down-rounds almost never happen here. The funds across the region are still relatively young. The LP relationships are new. Nobody wants to be the one writing a markdown into their portfolio, having that conversation, taking that medicine. So instead, they hold the valuation flat, keep the paper TVPI looking reasonable, and wait.

You can talk about your book value multiple all you want. If the company can’t raise and can’t grow, the number isn’t real.

The downstream problem: there are cases where this dynamic is actually blocking deals. An investor who doesn’t want to see a down-round may resist a transaction that would otherwise be good for the company, because accepting it means acknowledging the valuation they’ve been carrying is wrong.

Sometimes you have to take one step back to take two steps forward. That’s not a comfortable thing to do. But it’s more honest than pretending nothing is wrong until there are no options left.

* * *

SkyeChip and Malaysia’s Chip Moment

I want to start this one with genuine enthusiasm, because it deserves it.

SkyeChip Bhd lists on Bursa Malaysia’s Main Market on May 20th. The public tranche closed 95 times oversubscribed. Total retail demand hit RM 3.04 billion. The largest retail subscription in Malaysia since Petronas Chemicals in 2010, sixteen years ago.

The whole AI and chip investment wave has been impossible to ignore. NVIDIA’s share price trajectory. The compute boom. The data centre buildout. And now, emerging from Penang, a Malaysian company that sits right in the middle of that stack. That’s a big deal for this ecosystem.

Upfront caveat: I’m not a semiconductor expert. What follows is based on my research into the prospectus and what’s been circulating in the analyst and retail investor community. Take it in that spirit.

What SkyeChip Actually Does

Malaysia’s semiconductor sector has historically been dominated by the back end: assembly, testing, packaging. Important work. But it’s the low-margin end of the chain. The government has pushed for years, through NIMP 2030, through IC design parks in Selangor and Penang, through various national initiatives, to move the industry up the value chain into front-end design.

SkyeChip is the poster child for that ambition. It’s a fabless IC design company, it doesn’t manufacture chips, it designs silicon intellectual property. Reusable building blocks that chip makers integrate into their own products.

Think of it this way: TSMC makes the chips, NVIDIA designs what goes on them. SkyeChip is not saying they service either of those companies, but the analogy holds, they sell the blueprints for specific components that go inside chips. Their flagship IP is HBM3E: high-bandwidth memory interface technology, the memory architecture inside the AI accelerators that run the large language models powering frontier AI.

That’s the tie-in to the chip craze. And it’s why the hype is real. This isn’t fabricated. The technology is real.

The National Story

The government is leaning in hard, and in this case the support is substantive not just rhetorical. SkyeChip gets access to Arm Holdings design tokens through Malaysia’s Silicon Vision initiative, a national licensing arrangement that gives Malaysian companies access to Arm’s IP architecture. That’s a genuine strategic asset, not a marketing line.

The Deputy Minister attended the prospectus launch and talked about SkyeChip potentially reaching the level of Broadcom. Broadcom is a $700 billion company. SkyeChip is listing at RM 1.6 billion. The ambition is clear. The road is long.

But what matters is that this company is creating a visible proof point, that a Malaysian IC design house can be built, can reach a meaningful scale, can list on the main market, and can attract global attention. The next founder who wants to build something like this now has an example. That matters for the ecosystem in ways that go beyond the specific valuation.

The Numbers Worth Noting

Revenue more than doubled over two years. Profit margins around 30%. Analysts projecting roughly 31% earnings CAGR over three years, with the most bullish target price close to double the IPO price of RM 0.88.

The business model, IP licensing, is a proven high-margin, scalable model. Arm, Cadence, Synopsys. These are multi-billion dollar businesses built exactly this way: create the IP once, license it repeatedly. SkyeChip isn’t reinventing the model. It’s executing on it with new IP in a hot category.

The Risks That Deserve Honest Attention

China Revenue and US Export Controls

For the seven months ending October 2025, China accounted for 73.3% of revenue. Almost three-quarters of the company’s most recent revenue came from Chinese fabless IC companies selling advanced HPC and AI chips.

The prospectus explicitly acknowledges that if any of their customers are added to the US Entity List, supply must be suspended. None are listed today, but today is a snapshot, not a guarantee. The company is also planning to open US offices, which creates a real balancing act between serving Chinese customers and operating in a US regulatory environment that is actively tightening controls on exactly this category of IP.

The Tax Exemption Expired

This is the one I keep coming back to.

SkyeChip has been operating under a Pioneer Status tax exemption, effectively a 2.7% tax rate. That exemption expired September 9, 2025. They applied for renewal. As of the last published date in the prospectus, the renewal is still under review.

The IPO is priced at 44x FY2025 earnings. Those earnings use a 2.7% tax rate that no longer exists. Normalise to a standard 25% rate and you’re paying closer to 57x.

Most analysts will have noted this. But it’s worth being explicit about: the multiple headline is priced on a tax rate that hasn’t been legally valid for eight months and may not be renewed. That’s a material question sitting unresolved at the point of listing.

Revenue Quality and Customer Concentration

Top three customers represent around 60% of FY2025 revenue. More importantly, the revenue model is largely non-recurring, lump-sum contracts, one-off sales, high upfront. You need to keep winning new work to replace completed contracts.

Retail investors who have done deep dives on the prospectus, the i3investor and KLSE Screener community has been thorough here, have flagged that several of the largest customers from earlier years no longer appear as active. Replaced by new Chinese customers with sub-one-year relationships. Customer names are undisclosed so independent verification isn’t possible, but the pattern is worth understanding before you subscribe.

Where I Land

Malaysia needs stories like this. We need proof points that deep tech can be built here, that front-end design is achievable, that a Malaysian company can capture global demand in a critical technology category. SkyeChip creates that proof point. Congratulations to the team and their investors, genuinely.

The technology is real. The Arm access is real. The revenue growth is real. There’s genuine substance here and, looking at comparable companies globally, there’s still room for upside even from the IPO price.

The risks are also real. China concentration, an expired tax exemption, non-recurring revenue, some customer churn buried in the prospectus. None of these are necessarily deal-breakers. All of them require the optimistic scenario to hold.

Watch the listing day on May 20th. The market will be more honest than any analyst note about how much of the 95x was conviction and how much was leverage-financed retail applications planning a day-one flip.

* * *

OpenAI and Anthropic Just Told You the Hard Part

This is the most globally significant story of the week. And I think it has the most direct implication for founders building in Southeast Asia right now.

Within the same week, Anthropic first, then OpenAI, both companies announced they are building enterprise AI services companies. Not products. Not model updates. Not API pricing changes. Services companies. Engineers going inside client organisations and building AI systems for them.

What They Announced

OpenAI announced on May 11th. They’re calling it the OpenAI Deployment Company. Launching with over $4 billion in initial investment from 19 founding partners, TPG leading, with Bain Capital, Brookfield, Goldman Sachs, SoftBank, McKinsey, and Capgemini in the group. OpenAI also acquired Tomoro, an applied AI consulting firm, and brought roughly 150 engineers into the venture from day one. OpenAI retains majority ownership.

The model: Forward Deployed Engineers (FDEs) embedded directly inside client organisations. They work with business leaders and frontline teams to identify where AI can have the biggest impact, redesign workflows around it, and build production systems connected to the company’s actual data and infrastructure.

Anthropic announced a week earlier, backed by Blackstone, Hellman and Friedman, Goldman Sachs, General Atlantic, Apollo, GIC, and Sequoia. Same fundamental concept. Their framing specifically targets mid-market: community banks, mid-size manufacturers, regional health systems. Companies that could benefit enormously from AI but don’t have the internal resources to build and run frontier deployments.

When you look at the roster of investors across both of these efforts, you’re seeing a significant portion of global private capital touching large segments of the broader economy. This is not a side bet.

The Palantir Model

To understand why this matters, you need to understand what Palantir built over the last two decades.

Palantir’s entire model was built on one idea: you can’t sell complex software to complex organisations and expect them to use it well. You have to embed engineers inside the organisation. Work through the legacy systems, the internal politics, the messy reality of how things actually get done inside a large enterprise. Build something that functions in that specific environment.

That made Palantir extraordinarily sticky. Once you’ve had a team embedded inside an organisation for months, rebuilding core operational workflows around your platform, good luck ripping that out. The model is controversial. Critics call it consulting dressed as software. Believers say it’s the only honest way to sell software to organisations that don’t know what they need.

OpenAI and Anthropic are applying that same logic to AI. At scale. With billions behind it.

If the models were easy to deploy, these services arms would not need to exist. Full stop.

The Deployment Gap Is the Real Problem

Enterprise AI has a gap that doesn’t get enough honest discussion. The models work. Claude works. GPT works. The demos are genuinely impressive. But when companies try to deploy these systems into actual operations, into fifty-year-old legacy software, complicated permission structures, compliance requirements, and workflows that have developed organically over decades, the complexity is enormous.

The gap between “this model is impressive” and “this model is running reliably inside our organisation and measurably improving how we operate” is not a small gap. It is enormous. And closing it requires human expertise, people who understand the technology and the specific operational context of the organisation.

The fact that both labs are committing at this scale to closing that gap is an admission. Model quality is not the bottleneck anymore. Deployment is the bottleneck. And that reframes where value sits in the AI stack.

The Inversion of SaaS

Here’s a framing I’ve been thinking about. The SaaS era was defined by software being light on the surface, an interface you accessed yourself. The software sat on top of your workflow but you still had to do the work. Self-service by design.

What these services arms represent is something different. The model is going deep into the workflow, understanding it, rebuilding it, and then leaving behind something that runs with minimal human intervention. You’re not delivering software. You’re delivering a running operation. Services as software.

If that model sticks, and the fact that it’s being backed this heavily suggests it will, the companies that win are not the ones with the best model. They’re the ones who can deploy the best model inside the most complex environments, with the most contextual understanding of how those environments actually work.

What This Means for Southeast Asia

OpenAI’s Deployment Company is starting in US enterprise. Anthropic is starting in US mid-market. Neither of them is starting in Southeast Asia.

That means the deployment gap in this region is not going to be closed by Silicon Valley in the near term. Someone local has to do it.

The bank in KL running a fifty-year-old core banking system. The Indonesian manufacturer with warehouses of paper records. The healthcare group operating across five countries with different languages and different regulatory frameworks in each market. These aren’t problems that a foreign firm can parachute in and solve. They require local knowledge, local language, local relationships, and long-term on-the-ground presence.

The two most credible AI labs in the world just confirmed there is a structural, multi-billion dollar opportunity for exactly this business. The window to build it before the global players get here is not unlimited.

If you are building an AI services or implementation company in Southeast Asia right now, this week’s announcements are a green light. Pick up the pace. The clients will move slowly, that’s fine, enterprise always moves slowly. You move fast. Get embedded. Build the local relationships. Develop the deployment expertise. Because once you’re in and the workflows are built around what you’ve built, it becomes very hard to replace.

And one more signal worth noting: when AI labs start building services arms, it tells you something about the model layer. If being the best model was a durable, defensible moat, you would not need a services company. You would just keep making the model better and let it sell itself. Both companies have genuinely good models. They’re still building this.

The future isn’t won at the model layer. It’s won at the integration layer, the workflow layer, the trust layer. For founders building AI companies in Southeast Asia, that’s the competition you’re actually in. And it’s a winnable one.

* * *

What These Three Stories Say Together

Put them next to each other and they’re telling one thing.

The funding market is leaner than the headlines suggest. Capital is concentrating, in Singapore, in AI, in infrastructure. The zombie problem is real and growing quietly. There are silver linings: Malaysia is moving, agentic AI has consistent demand, and the data centre boom is real even if it distorts the quarterly numbers.

SkyeChip is the most tangible proof point this ecosystem has produced in years that big, globally relevant deep tech can come out of Malaysia. Whether it becomes a durable business depends on questions the prospectus cannot yet answer. The execution has to turn the IPO moment into something lasting.

And the global AI labs just spent billions telling you that the hard part of AI isn’t the model. In Southeast Asia, the opportunity to do that hard part, the deployment, the integration, the on-the-ground expertise, is wide open and freshly validated.

The founders who understand that and move on it in the next twelve to eighteen months are the ones worth watching.

This post accompanies the SEA of Startups episode for the week of May 13, 2026. Listen wherever you get your podcasts.

Real. Raw. Relatable.

SEA of Startups | Kevin Brockland

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

This week’s episode is a news episode. No guests. Just four stories that I think every founder, investor, and operator in Southeast Asia should be paying attention to right now.

Here’s what we cover, and why each one matters.

1. China forced Meta to unwind a completed acquisition. Mid-honeymoon.

In December, Meta acquired Manus — the AI agent startup that went viral in 2025 as China’s answer to deep research tools. The deal closed. Manus’s website was already saying it was part of Meta.

On April 28th, Beijing’s NDRC told both parties to reverse it.

The Singapore-washing playbook — where Chinese founders restructure as Singapore entities to access US capital — is now provably dead. Beijing just proved it can reach into a completed acquisition, across jurisdictions, and pull the plug.

But the surface story is not the interesting story. The interesting story is the mechanics of what an “unwind” actually looks like. Money has already flowed through to investors and their LPs. Engineers have been working inside Meta for weeks. Knowledge transfer has happened. How do you reverse that?

And then there’s the Meta question. Did they make a mistake — or did they knowingly race the regulator, betting that if they got the technology embedded before enforcement could land, a slow unwind would be better than no acquisition? Their public statement — “the transaction complied fully with applicable law, we anticipate an appropriate resolution” — says absolutely nothing. Which might be exactly the point.

Singapore has been conspicuously silent throughout all of this. What that silence costs them is a conversation the episode goes deeper on.

2. eFishery. Nine years. And it still doesn’t feel like enough.

Gibran Huzaifah was sentenced to nine years on April 29th. Two other former executives received nine and seven years respectively.

The numbers, if you haven’t heard them: the company told investors it generated $752 million in revenue from January to September 2024. Actual revenue was $157 million. They reported a $16 million profit. The actual result was a $35 million loss.

SoftBank. Temasek. KWAP — Malaysia’s civil servant pension fund. All recovering less than ten cents on the dollar.

But this episode is not a crime recap. The eFishery story is a prompt for a harder question about what kind of ecosystem we’re building here.

Fraud exists on a spectrum. At one end: criminal fabrication at scale. At the other: things that happen every week across the region that would never see a courtroom — vanity metrics dressed as traction, pilots treated as revenue, LOIs presented as signed contracts. None of that is eFishery. But it is on the same continuum.

And it is not only founders. Investors do it too.

The reason this matters beyond the immediate case is economic. In a high-uncertainty market like Southeast Asia, trust is the operating system. When it erodes — when every investor assumes every founder is telling the most optimistic version of the truth — the whole system gets more expensive. More friction. More time on verification. Fewer deals done.

A high-integrity environment is a high-output environment. The ecosystem gets the standards it is willing to enforce.

3. Indonesia capped ride-hailing commissions at 8%. GoTo just posted its first-ever profit. Congratulations.

On May 1st — International Workers’ Day, timing very much intentional — President Prabowo signed a regulation capping the maximum commission ride-hailing platforms can take from drivers at 8%. Down from 20%. Drivers now get a minimum of 92% of every fare.

GoTo shares dropped nearly 6% on the news. Analysts estimated the ride-hailing segment accounted for roughly 48% of GoTo’s EBITDA. Grab, which derives about 20% of its total EBITDA from Indonesia, is also in the firing line.

Both companies will either raise fares, eat the margin hit, or some combination of both. None of those options is clean.

Here is the part that might be unpopular in a room full of investors: Prabowo is not entirely wrong.

Indonesia has around four million ride-hailing drivers. The platform without the driver is just an app with nowhere to go. The economics for drivers have been genuinely rough. The system was designed to extract maximum value from a class of workers with very little negotiating power.

The underlying question — how do we ensure the people who actually do the work get a fair share of what they create — is legitimate. If platforms do not answer it voluntarily, governments will answer it for them.

The risk, of course, is that fares go up, volumes drop, and drivers end up worse off than before. That is the irony of heavy-handed regulation. But that is a problem for GoTo and Grab to solve. They had the data. They should have got ahead of this before a president had to sign a decree on Workers’ Day.

4. Malaysia is building gas plants to power AI data centres. The energy transition did not plan for this.

This week, a Melaka-based company called DPS Resources — until recently primarily a furniture and property developer — announced it signed an MOU with an Alibaba affiliate to explore building a $1.1 billion AGI data centre in Melaka. 150 to 180 megawatts. DPS provides the land, the power, the infrastructure. Alibaba’s entity handles operations and brings the computing demand.

This deal is not an anomaly. It is a perfect emblem of what is happening across Malaysia right now. Everyone wants a piece of the data centre gold rush. The question not being asked loudly enough is whether Malaysia actually has the power to sustain it.

TNB’s pipeline is 7,500MW across 56 data centre projects. Current actual load from those facilities: 850MW. The draw-down is coming as facilities rack up through 2026. At the same time, 6,400MW of coal-fired generation is scheduled for retirement between 2029 and 2031.

To cover those retirements and meet rising demand, Malaysia needs roughly 12,000MW of new generation by 2031.

Right now, the Energy Commission has an open tender — NewGen26 — for new gas-fired generation to plug that gap. Bids close July 1st. Eight weeks away. This is Malaysia racing to build baseload capacity before the demand wall hits. The fact that it is gas, not solar, tells you everything about the timeline pressure.

The Iran conflict makes this personal. TNB’s Automatic Fuel Adjustment mechanism means global oil and gas price spikes feed directly into Malaysian electricity bills within 30 days. Data centres in Johor were approved on the premise of cheap, stable Malaysian electricity. That premise is now under pressure from a war on the other side of the world.

The deeper question is who actually benefits from this boom. DPS provides the land and the power. Alibaba keeps the data, the models, and the IP. Research consistently shows data centres create the lowest number of jobs per square foot of any major facility type. Thousands of construction roles during the build, then roughly 200 operational staff when running.

Malaysia is providing the real estate, the utilities, and the environmental cost. The hyperscalers are keeping the value.

That is not a reason to stop. But it is a reason to be far more deliberate about what we are trading and what we are getting in return.

Watch the episode

Four stories. One theme running underneath all of them: the rules are being rewritten. Who controls AI. Who controls capital flows. Who gets a fair share of the value created. Who owns the infrastructure the future runs on.

These are not settled questions. They are live negotiations — between governments, between companies, between regions.

Southeast Asia is not a passive observer in any of this.

[Watch / listen to the full episode → link]

SEA of Startups is a podcast for founders, investors, and operators building in Southeast Asia. Real. Raw. Relatable.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

Episode Title: The New Space Age Is Actually Here | Artemis II, SpaceX IPO & The Rise of Orbital Infrastructure

Episode Summary

Right now, four humans are flying around the moon. Not in a simulation. Not in a film. For real. Kevin uses the launch of Artemis II on April 1, 2026 as the jumping-off point for a deep dive into the most consequential shift in space exploration since the Apollo era — and why this time, it's not just governments leading the charge.

From SpaceX's against-all-odds origin story to the trillion-dollar IPO that just rocked public markets, this episode charts how the economics of space fundamentally changed, what that means for a new generation of startups, and whether the science fiction stories we grew up watching are finally, actually, coming true.

What We Cover

Artemis II — Who's on board, what they're testing, and why this 10-day lunar flyby matters beyond the symbolism

The cost collapse — How SpaceX drove launch costs from $10,000–$20,000/kg down to under $2,000/kg (and potentially below $100 with Starship)

The space economy by the numbers — $8B+ raised in 2025 alone, 154% YoY growth, 35,000+ companies globally, a projected $1T market by 2033

Startups reshaping the supply chain — Rocket Lab, Apex, Hadrian, The Exploration Company, and the infrastructure plays most people aren't watching

Earth observation goes commercial — How Planet Labs and others turned satellite data into a sovereign government revenue model

The SpaceX IPO — Filed confidentially the same day as Artemis II, targeting a June NASDAQ listing at a reported $1.5–2T+ valuation (potentially the largest IPO in history)

Starlink's numbers — 10M subscribers, $10B revenue in 2025, projected $24B by end of 2026, and what direct-to-cell really means

Orbital data centers — Star Cloud's H100 GPU satellite, Google's Project Suncatcher, Blue Origin's TeraWave, and why AI's energy problem might get solved in orbit

The moon as infrastructure — Lunar ice mining, the South Pole fuel depot play, and Lone Star Data Holdings building a data center on the lunar surface

The sci-fi question — Are the stories we grew up with finally coming true?

Key Numbers

StatFigureSpace tech funding raised in 2025$8B+YoY growth in space funding154%Projected space market by 2033~$1 trillionNew employees added in the past year~200,000Cost to orbit in the 1990s$10,000–$20,000/kgCost to orbit today (Falcon 9)Under $2,000/kgStarlink subscribers (end of 2025)10 millionStarlink revenue 2025$10BSpaceX IPO reported valuation$1.5–2T+Star Cloud Series A valuation$1.1B (18 months old)

Companies & Missions Mentioned

SpaceX · Artemis II / NASA · Rocket Lab · Planet Labs · Apex · Hadrian · The Exploration Company · Star Cloud · Lone Star Data Holdings · Blue Origin (TeraWave) · Google (Project Suncatcher) · xAI · Starlink

People Mentioned

Reed Wiseman — Artemis II Commander

Victor Glover — Artemis II Pilot; first Black person to travel to the moon

Christina Koch — First woman to travel to the moon

Jeremy Hansen — First Canadian to travel this far from Earth

Jared Isaacman — NASA Administrator

Elon Musk — SpaceX / xAI / X

Chad Anderson — Founder, Space Capital

Quotes Worth Sharing

"SpaceX didn't just build a business. It rewrote what was possible."

"The interplanetary story is no longer confined to Elon Musk's conference slide decks. It's in regulatory filings. It's in rocket test programs. It's in the hiring plans of hundreds of companies."

"The gap between what the stories promised and what actually happened at times felt like a wound. But now I look at what's actually happening and I find myself genuinely surprised."

Follow the Show

🎙️ SEA of Startups — Real. Raw. Relatable. YouTube | TikTok | Instagram

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

Most AI content gives you a framework. Tiwa York gives you a verdict.

The founder who built Kaidee to 35 million users and guided it to a successful exit sat down with SEA of Startups and said what most operators are afraid to say out loud: your team is probably performing AI adoption, not doing it. And the longer you stay there, the harder it gets to move.

Here’s what he actually said — the numbers, the examples, the provocations.

The 5 Levels of AI Maturity (And Why 1.5 Is a Trap)

Tiwa’s framework runs from 0 to 4. Most conversations stop at listing the levels. The more important conversation is why so many companies get stuck halfway through Level 1.

Level 0 — Unaware: No AI tools in use. Working like it’s 2019.

Level 1 — Curious: ChatGPT is bookmarked. It gets used for emails and translation. Actual work output: unchanged.

Level 1.5 — The Trap: This is where Tiwa spends most of his time on stage. A few people are experimenting. Strategy decks mention AI. But workflows, decisions, and output haven’t moved. He calls this adoption theater — and it’s where the majority of SEA companies currently sit.

Level 2 — Active: AI is genuinely built into daily work. Measurable productivity gains of 25–50%.

Level 3 — Integrated: Multiple AI tools connected in smooth workflows. The data analyst goes from one report a week to one a day. The PM tests ideas overnight with simulated customers. 2–3x productivity — and completely redesigned ways of working.

Level 4 — Transformative: Creating value streams that simply didn’t exist before. Tiwa estimates this is roughly 2% of the global workforce today.

The goal isn’t to inch from 1.5 to 2. It’s to move from 1.5 to 3, and then to 4. Anything less is rearranging deck chairs.

The Mental Model That Changes Everything

Tiwa’s most useful reframe isn’t a framework — it’s a metaphor.

Think of AI as the most capable but most forgetful intern you’ve ever hired. It can do almost anything better than any employee on your team. But the moment it leaves a conversation, it remembers nothing. Zero context. Starting from scratch.

This metaphor matters because it tells you exactly what your job is: you’re not a user of AI. You’re a systems designer for AI. Your task is building the handoff infrastructure — the context-carrying mechanisms, the memory systems, the structured prompts — that prevent that amnesia from killing your output quality.

Tiwa draws a direct parallel to the Toyota Production System. You’re not optimising one conversation. You’re building a manufacturing process for intelligence, with daily standups, continuous improvement loops, and institutional memory that compounds over time.

Most companies treat AI like a vending machine. High performers treat it like a factory floor.

The Numbers That Should Stop You Mid-Sentence

If you think the efficiency gap between good and great AI usage is somewhere between 20–30%, Tiwa has a number for you.

The difference between a 30% productivity gain and a 300x productivity gain isn’t the model you’re using. It’s how you’re using it.

That’s not a typo. 300x. The delta between someone using AI as a faster search engine and someone who has built genuine fluency — with context management, iteration discipline, and system-level thinking — is not incremental. It’s categorical.

On token economics specifically, Kevin cited Jensen Huang’s framing directly: a developer earning $500K annually should be spending roughly $250K a year in AI tokens. That’s the ratio of a high-performance AI-native engineer. For context: serious power users are already spending $500+/month on tokens. Some AI-native startups are at $1,000 per person per day.

If your developers aren’t asking for AI budget, Tiwa’s take is unambiguous: that’s a performance issue.

The Hiring Freeze Argument (And Why It’s Not Crazy)

The most provocative position Tiwa took in the recording:

Freeze all hiring until your AI implementation is complete.

The reasoning is mathematical. Communication pathways explode non-linearly with headcount:

* 5 people → 10 pathways

* 10 people → 45 pathways

* 20 people → 190 pathways

Every person you add before you’ve stabilised your AI workflows creates coordination overhead that compounds. You’re layering human complexity on top of unresolved process complexity. The problems don’t add — they multiply.

The implication for most early-stage SEA founders: your instinct to hire for growth may be the thing slowing your growth. A team of 6 people who are genuinely at Level 3 will outrun a team of 15 people stuck at Level 1.5, every time.

The Middleware Trap: A Warning for Builders

Tiwa is an investor. He’s pattern-matching on where value will be captured — and where it will evaporate.

His verdict on horizontal and middleware AI companies: 18-month obsolescence risk. The major frontier models are absorbing middleware functionality as a matter of course. If your moat is sitting between the model and the enterprise, that’s a shrinking gap.

The defensible positions he sees in SEA:

* Vertical solutions with deep workflow integration and hard-to-replicate domain understanding

* Regulated, complex legacy environments where switching costs are real and proprietary data is locked in

* Physical AI — Tiwa cited MUI Robotics, which has deployed an AI tongue (taste and smell sensors) across dairy companies, water utilities, and hotel renovation monitoring, and is currently running a research project on early liver cancer detection through smell. 300+ clients. 50+ multinationals. That’s not a middleware play.

The common thread: proprietary data, physical integration, or regulatory complexity. If you can be replaced by a model update, you’re not building a business — you’re building a feature.

Two Real Examples, Not Hypothetical Ones

The Jira/Confluence Replacement: A software development house replaced its entire project management stack — Jira, Confluence, the lot — in four days using AI-assisted development. Annual savings: $24,000. More importantly, they own the system now. No vendor dependency. No per-seat pricing. No waiting for a roadmap that doesn’t match their workflow.

The HubSpot Replacement: A friend of Tiwa’s replaced their entire HubSpot instance with a custom-built CRM in eight hours of AI-assisted coding. Eight hours. The off-the-shelf tool cost thousands annually and didn’t fit the workflow. The custom solution does — and it cost a weekend.

The pattern here isn’t “build vs. buy.” It’s “stop buying things that make you dependent when you could own the thing in a day.”

What AI-First Actually Requires From Leadership

Tiwa’s framework for leaders isn’t about tool selection. It’s about accountability architecture.

The key shifts:

Every function owns its own transformation. This can’t live with the CTO alone. Engineering, product, marketing, finance, customer success — every team lead is responsible for their own AI integration roadmap.

Model the behaviour publicly. If leadership isn’t visibly using AI — and visibly failing with it, learning from it, sharing what they found — no one else will take the cultural signal seriously.

Measure outcomes, not activity. Logins aren’t fluency. Licenses aren’t execution. The metrics that matter: workflow velocity, decision speed, output quality. Not hours of AI training completed.

Daily continuous improvement. Not a quarterly AI review. A daily standup cadence for what’s working, what broke, what gets refined tomorrow. Toyota didn’t build the production system in a sprint. Neither will you.

The Real Question

Tiwa closed with the line that stayed with everyone in the room.

“The question isn’t how do we find extraordinary people. It’s whether extraordinary people get unleashed inside this org — or leave to do it on their own.”

For founders in SEA: you probably already have the talent. The judgment is in the building. The only variable is whether you build the systems that let it operate at full power — or whether you stay at Level 1.5 long enough that the people who figured it out first come back to compete with you.

Watch the full conversation with Tiwa York on SEA of Startups

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

In the ever-changing landscape of technology and business, the term “SaaSpocalypse” has emerged to describe the recent downturn in public software stocks. But what does this mean for the future of SaaS companies, especially in Southeast Asia? In this blog post, we’ll explore the nuances of the SaaSpocalypse, the potential for growth amidst disruption, and what established and emerging companies can do to adapt.

Understanding the SaaSpocalypse

The term SaaSpocalypse refers to the recent significant decline in the valuations of publicly traded SaaS companies. This decline has raised concerns about the future viability of these companies. But is the doom and gloom justified?

The Current Landscape

- Valuation Adjustments: Many SaaS companies have seen their valuations drop sharply, leading to discussions about overvaluation in the sector. As Chris Birrell notes, some of these companies were indeed due for a correction.

- Growth Continues: Despite the downturn, many SaaS companies are still experiencing growth rates of 15-20% year-over-year, which, although lower than previous highs, indicates resilience in the market.

Key Insight: The SaaS market is not dying; it’s evolving. Companies that can adapt to new technologies, especially AI, may find new opportunities for growth.

The Role of AI in SaaS

AI is a game-changer for many industries, and SaaS is no exception. As the demand for AI integration grows, traditional SaaS companies must adapt.

Embracing AI Technologies

- Increased Demand for AI Solutions: Companies are under pressure to integrate AI into their workflows. This presents both a challenge and an opportunity for incumbents who can leverage their existing customer relationships to offer new, AI-driven solutions.

- The Risk of Disruption: While established companies may have a strong foothold, they are not immune to disruption. New entrants who can offer innovative solutions may quickly gain traction.

Example: Companies like Salesforce are well-positioned to sell AI-driven solutions, thanks to their existing customer base and established workflows.

Navigating Change: Strategies for SaaS Companies

As the industry evolves, SaaS companies in Southeast Asia must consider their strategies carefully. Here are a few key areas to focus on:

Focus on Core Competencies

- **Defensible Moats**: Companies with deep integrations into their clients’ workflows are better positioned to weather market fluctuations. Understanding what makes your service indispensable can help you maintain customer loyalty.

- **Avoiding the Surface-Level Solutions**: Companies that offer point solutions without deep integration risk losing market share to more comprehensive platforms.

Capitalizing on Regional Nuances

Southeast Asia is a unique market, and understanding local dynamics can provide a competitive edge.

- Local Expertise: Companies with founders who understand regional challenges are likely to succeed where larger, global firms may falter. This localized approach can help companies tailor their solutions to meet specific market needs.

The Future of SaaS in Southeast Asia

Looking ahead, what does the future hold for SaaS companies in Southeast Asia?

Opportunities Amidst Challenges

- Emerging Startups: As Chris mentions, startups that can build reusable software components tailored for AI-driven environments may find success. There’s a growing need for specialized solutions that can integrate seamlessly with existing workflows.

- BPO Evolution: Business Process Outsourcing (BPO) companies are also on the brink of transformation. By leveraging AI, they can enhance their service offerings and improve efficiency, setting the stage for a new era in service delivery.

Conclusion: Adapting for Success

In conclusion, while the SaaSpocalypse presents challenges, it also opens up avenues for growth and innovation. Companies that can adapt to the changing landscape—embracing AI, focusing on core competencies, and understanding regional market nuances—will be well-positioned to thrive in the future.

Key Takeaways:

- The SaaSpocalypse is not the end, but a transition.

- Embrace AI and focus on integration to maintain your market position.

- Understand regional dynamics to tailor your solutions for success.

---

Frequently Asked Questions

What is the SaaSpocalypse?

The SaaSpocalypse refers to the significant decline in valuations of publicly traded SaaS companies, raising concerns about the future of the industry.

How can SaaS companies adapt to the changing landscape?

By integrating AI solutions, focusing on their core competencies, and understanding regional market dynamics, SaaS companies can navigate the challenges ahead.

Is the SaaS industry dying?

No, the SaaS industry is evolving. Companies that can innovate and adapt will continue to thrive.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

Keywords

Meta, Manus, acquisition, Singapore, AI, geopolitics, startups, tech industry, business growth, investment

Summary

In this conversation, Kevin and Kim discuss Meta's recent acquisition of Manus, a Singapore-based startup, exploring its implications for founders in the region, the geopolitical landscape, and the evolving nature of AI in business. They analyze the rapid growth of Manus, the significance of Singapore as a tech hub, and the challenges posed by regulatory scrutiny. The discussion highlights the potential for Southeast Asia to emerge as a key player in the global tech ecosystem, while also addressing the complexities of company nationality and the future of AI amidst geopolitical tensions.

Takeaways

Meta's acquisition of Manus raises questions about the future of startups in Southeast Asia.

The deal signifies a shift in how tech companies navigate geopolitical landscapes.

Manus's rapid growth showcases the potential for startups in the region.

Acquisitions are not just about money; they often buy time and talent.

AI is changing the valuation landscape for tech companies.

Singapore is becoming a strategic hub for tech companies looking to scale globally.

The concept of 'Singapore washing' raises important questions about company nationality.

Geopolitical tensions could impact future tech acquisitions.

The success of Manus could inspire more founders in Southeast Asia.

Southeast Asia has the potential to be a significant player in the global tech ecosystem.

Titles

Meta's Bold Move: What It Means for Founders

Navigating Geopolitics in Tech Acquisitions

Sound bites

"They just bought time."

"Does it really matter? Not really."

"Singapore is the neutral zone."

Chapters

00:00 The AI Landscape and Major Players

02:45 Geopolitical Implications of AI Investments

05:53 The Role of Singapore in the Global Tech Ecosystem

08:54 The Evolution of AI and Market Dynamics

11:54 Regulatory Challenges and Market Valuations

14:17 The Future of AI and Founders' Perspectives

18:01 Navigating Nationality and Compliance in Tech

20:45 The Balance of Speed and Long-term Value Creation

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com

Summary

In this conversation, Kevin and Kim discuss the recent merger of Elon Musk's companies, particularly focusing on the implications of combining AI and space technologies. They explore the potential of data centers in space, the evolving role of Tesla, and the regulatory challenges that come with these advancements. The discussion also touches on the future of sovereignty in space and the messy landscape of regulations that may arise as private companies take a more significant role in space exploration.

Takeaways

Elon Musk is merging his companies to simplify operations.

The merger signifies a shift towards a unified intelligence layer.

Data centers in space could revolutionize computing.

Tesla's role is evolving beyond just electric vehicles.

Regulatory challenges will complicate space exploration.

Sovereignty in space is a complex issue.

The landscape of space regulations is becoming messy.

Private companies will play a crucial role in space.

Non-terrestrial data centers are on the horizon.

The future of AI is tied to its infrastructure location.

Titles

The End of the Discrete Company Era

Merging AI and Space: A New Frontier

Sound bites

"AI just got X'd."

"Tesla isn't an EV company anymore."

"It's going to be messy."

Chapters

00:00 The End of the Discrete Company Era

02:07 The Merging of Tech Giants

05:48 Data Centers in Space: A New Frontier

09:53 The Unified Intelligence Layer

14:56 The Future of AI and Space Exploration

20:05 Regulatory Challenges in Space

24:54 The Wild West of Space Law

29:55 The Dawn of a New Era

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit seaofstartups.substack.com