Over the past few years, the conversation around startup funding has started to shift. Not in a dramatic way, but in how capital is actually distributed across stages.

For a long time, venture capital sat at the centre of everything. Founders built it with it in mind from day one. The path felt clear. Raise early, scale fast, move through rounds. That model still exists, and venture capital continues to play a major role globally. Strong companies are still being funded, large rounds are still happening, and the system itself remains active.

But the shape of that system has changed.

Globally, venture capital today tends to engage more decisively when there is clearer traction. Businesses that show stronger execution, more defined markets, and visible growth pathways are where attention concentrates more naturally. This does not reduce the presence of venture capital. It shifts how and when it becomes most relevant.

In Southeast Asia, this pattern is even more visible. The ecosystem has matured, and with that comes a clearer separation between stages. Venture capital is still very much part of the journey, but it tends to play a more central role as companies move into growth phases, where there is already a base to build on.

At the earlier stages, the picture looks different.

Early-stage capital is still active, but it no longer sits in one place. It is distributed across a wider set of channels. Accelerators, incubators, venture studios, corporate programs, and a growing layer of grant and hybrid funding systems all play a role in supporting new ventures. In many cases, this is where early momentum is built.

These channels are not simply smaller versions of venture capital. They operate with different objectives. Some are designed to support ecosystem development, others focus on specific sectors, and many are aligned with institutional or policy-driven goals. The capital is there, but it sits within frameworks that are structured differently.

From the outside, this creates a sense that there is more capital available across more sectors than before. In many ways, that is true. But it also means that accessing this capital requires a clearer understanding of how each part of the system works.

A founder today may engage with multiple funding pathways at the same time. Applying to an accelerator, exploring grant opportunities, speaking to early-stage investors, and preparing for future venture rounds. The intent is to increase the chances of funding. The effort is broad.

What often remains unchanged is the approach.

Also Read: Why inclusive hiring matters for a startup ecosystem

The same narrative, the same materials, and the same assumptions are carried across different funding channels. A pitch designed for venture capital is used in a grant application. A grant proposal is framed with a venture-style story. Adjustments are made, but the core remains largely the same.

When outcomes do not change, it becomes difficult to interpret. It begins to feel like a reflection of the venture itself.

In reality, each funding channel operates as a filter before it becomes an opportunity. The evaluation begins with alignment. Stage, geography, structure, use of funds, and intent all shape how an application is read. These are not always visible from the outside, but they influence outcomes early in the process.

If the alignment is off, the process rarely moves forward in a meaningful way.

Because these filters are not always clearly communicated, founders tend to focus on what they can control. The pitch is refined, the message is sharpened, the projections are updated. Then the process repeats, often across multiple programs.

Over time, this creates a pattern where effort increases, but outcomes remain similar.

Seen from a broader perspective, this is less about the absence of capital and more about how that capital is structured.

Venture capital continues to play a central role, particularly as companies move into stages where scale and growth become more visible. At the same time, early-stage capital is active across programs, institutional channels, and sector-focused funding systems. Each layer is functioning, but each operates differently.

The result is a funding landscape that is more layered than before.

Also Read: Unlock your enterprise agility to unleash the potential of your startup

For founders, this means that the journey is no longer about approaching a single source of capital. It is about understanding how different forms of capital connect to different stages of the business. Moving through these layers in a way that matches the venture’s current position becomes increasingly important.

Without that alignment, the process can feel repetitive. Applications go out, responses come back, and the cycle continues without clear movement. With it, the same landscape begins to make more sense.

The capital is there. In many ways, there is more of it across more channels than before.

The difference lies in how it is approached and how well each venture fits within the system it is entering.

Understanding where startup money is really coming from today is not just about identifying new sources. It is about recognising that each source operates on its own logic, and aligning with that logic early in the journey.

Because in today’s funding landscape, there is no shortage of capital conversations. Only a shortage of alignment.

And if everything still feels confusing, there is always the classic fallback. 
Update the pitch deck, change the font, add a bigger market size slide, and try again.

Sometimes that works.

Just not for the reasons most people think.

So before doing that, it is worth stepping back and looking across the full spectrum of capital available today, venture capital, grants, blended finance, catalytic capital, and everything in between, and deciding where the real fit actually is.

—

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 post Where startup money is really coming from today appeared first on e27.

The Philippines’s services sector is indisputably the country’s engine of growth and employment. Today, it accounts for more than half of the national GDP and has posted growth rates that, at times, rival those of fast‑growing economies in the region.

Yet beneath headline growth lies a structural challenge: much of the expansion has been concentrated in labour-intensive, low‑productivity activities—traditional retail outlets, basic hospitality, and standardised business process outsourcing (BPO) roles. These activities provide essential jobs and incomes, but they lack the capital intensity, technological sophistication, and organisational complexity necessary for sustained gains in output per worker.

Also Read: The hidden tax on Philippine SMEs: Unreliable infrastructure

From a “Services Growth Problem” to a “Services Model Problem”

As per the Philippine Private Capital Report 2026 by Foxmont Capital Partners, this is less a services growth problem and more a services model problem. Persisting with expansion primarily through increased headcount in low‑productivity roles risks trapping the Philippines at middle‑income levels, aka the middle‑income trap. Breaking through requires reimagining how services are delivered: shifting toward knowledge‑based, system‑driven, and technology‑enabled models that raise productivity per worker and increase the share of value captured domestically.

Why the distinction matters:

• Employment expansion absorbs millions of new labour market entrants and supports livelihoods across regions.
• Productivity growth determines sustainable improvements in living standards and the country’s fiscal capacity to finance public services.

Retail: Platformisation, logistics, and the sari‑sari store opportunity

Traditional retail (wet markets, sari‑sari stores, and neighbourhood shops) remains central to Philippine commerce, particularly in provincial and low‑income urban communities. These formats are low‑capital, labour‑intensive, and typically show low output per worker.

The rise of e‑commerce is beginning to decouple retail growth from physical store expansion. Platform ecosystems such as Shopee and Lazada have catalysed productivity gains through data‑driven merchandising, dynamic pricing, and logistics optimisation. While platform giants report very high productivity per worker in other markets, the Philippines can capture similar qualitative benefits by adapting models to local contexts.

Policy and business levers for retail transformation:

• Expand affordable broadband and mobile connectivity to shrink the digital divide between urban and rural areas.
• Improve last‑mile logistics: invest in local sorting hubs, cold chain for perishables, and rural delivery partnerships.
• Upskill micro‑merchants: provide simple digital tools and training so sari‑sari owners can adopt digital payments, inventory apps, and online listings.
• Strengthen financial services: micro‑credit and invoice financing enable merchants to stock higher‑value inventory and smooth cash flow.
• Ensure platform governance: enforce consumer protection, fair competition, and data privacy while encouraging local innovation.

IT‑BPM: Moving up the value chain with AI and analytics

The Philippines has built a world‑class IT‑BPM industry centred on voice‑based contact centres and transactional back‑office work. The sector has been a major export earner and employer in Metro Manila, Cebu, Clark, and other hubs. Yet automation threatens routine tasks, and scaling by sheer headcount offers diminishing returns.

Also Read: Philippines’s productivity problem starts in the classroom

The next wave of growth lies in higher‑value segments: analytics, knowledge process outsourcing (KPO), engineering services, digital transformation consulting, and AI‑enabled operations. These segments raise output per worker by embedding automation, natural language processing, and predictive analytics into workflows—reducing handling time, increasing first‑contact resolution, and offering advisory services that command higher fees.

Actions to accelerate IT‑BPM upgrading:

• Reform education: strengthen university and technical curricula in data science, software engineering, and domain specialisations (healthcare, fintech, legal process outsourcing).
• Boost lifelong learning: subsidise bootcamps, certifications, and company‑sponsored reskilling programs for mid‑career workers.
• Incentivise high‑value investment: targeted tax and non‑tax incentives for companies establishing analytics, R&D, and innovation centres in the Philippines.
• Promote local scaleups: expand venture capital and growth financing for SaaS firms and analytics platforms that can export services regionally.

Policy and institutional foundations

Transformation requires coordinated public‑private action across infrastructure, education, finance, regulation, and regional development.

Key levers include:

• Digital infrastructure: accelerate nationwide fibre builds, tower deployment, and spectrum allocation to make high‑quality connectivity ubiquitous.
• Education and training: integrate computational thinking and digital literacy into K–12, expand technical-vocational education, and scale industry‑aligned short courses.
• Financial inclusion: deepen digital payments, merchant lending, and credit scoring using alternative data to unlock SME investments in productivity tools.
• Regulation and trust: strengthen data protection laws and regulatory clarity for cloud services and cross‑border data flows to attract enterprise customers.
• Regional diversification: create incentives and shared infrastructure so higher‑value service hubs develop in Cebu, Davao, Iloilo, Clark, and other cities—reducing congestion in Metro Manila and tapping local talent pools.

Social considerations and inclusive transition

Technology and productivity gains can displace certain roles even as they create higher‑value ones. Managing this transition requires active labour market policies:

Also Read: Philippines’s quiet AI revolution is about work, not tech

• Reskilling and upskilling programs targeted at workers most likely to be affected by automation.
• Transition support and portable social protections (e.g., unemployment insurance, wage subsidies during retraining).
• Community outreach and accessible training channels—mobile training units, local government partnerships, and employer commitments to hire trained workers.
• Inclusive growth also means bringing women and underrepresented groups into higher‑value services through scholarships, targeted training, and workplace flexibility measures.

Realistic timelines and the Philippines’s comparative advantages

The shift from labour‑intensive to knowledge‑ and tech‑driven services is multi‑year but achievable within a decade with concerted effort. The Philippines has several strengths to build on:

1. A large, English‑proficient, young workforce.
2. An established BPO ecosystem and global reputation for service delivery.
3. Robust remittance inflows support domestic demand and entrepreneurship.
4. Growing digital consumer adoption and a vibrant startup scene.
5. Targeted policy, industry leadership, and investments in human capital can move the country from employment‑centric growth to productivity‑centric expansion.

Conclusion: Services that create value at scale

The Philippines’s services sector can remain a major employer while becoming a source of higher productivity and sustainable economic growth. The pathway out of the middle‑income trap is not merely more jobs, but better, higher‑value jobs that scale through knowledge, systems, and technology rather than headcount alone.

Also Read: From assembly line to innovation engine: Can Philippines climb the chip value chain?

When retail becomes platform‑enabled and IT‑BPM evolves into analytics and AI‑led services, output per worker rises, wages grow, and the economy creates more value with less proportional labour input—an outcome that benefits workers, firms, and the nation.

The post The productivity pivot the Philippines can’t delay appeared first on e27.

Thailand’s cybersecurity ecosystem expanded steadily through 2025, propelled by rapid digital transformation, surging cloud adoption, and stepped‑up investment in national data infrastructure.

Progress, however, has not been matched by commensurate operational depth. Ransomware campaigns have become more surgical, targeted attacks on financial institutions have intensified, and advanced persistent threats are showing up with greater frequency.

Also Read: Cyber risk is moving upstream but we’re still defending downstream

The result is a market that has built institutional scaffolding but still struggles to convert policy and procurement into resilient, repeatable defence.

A tense balance: Institutional gains vs operational shortfalls

Regulatory tightening and government digital initiatives have improved the country’s cybersecurity posture on paper. Data‑protection frameworks and cross‑agency programmes encourage firms to formalise security strategies. Yet intentions rarely substitute for capability.

Many organisations lack the in‑house talent and mature processes needed for effective detection, containment, and recovery. Demand for cybersecurity outpaces the available local expertise, prompting firms to rely heavily on managed security providers or fragmented point solutions.

This is not just a resource problem; it is a structural one. Budgeted security projects often deliver tactical improvements, but long‑term execution — continuous monitoring, threat hunting, architecture rework, and secure cloud migration — remains uneven. Smaller enterprises are particularly exposed: limited budgets and low security maturity make them easy targets and weak links in supply chains.

Threat landscape: Smarter adversaries, wider targets

The past year saw attackers shift from opportunistic intrusions to targeted, multi‑stage campaigns. Ransomware groups increasingly deploy double extortion tactics — encrypting data and threatening public leaks — amplifying reputational and regulatory risk. Financial institutions, with complex third‑party ecosystems and cloud dependencies, have been high‑value targets for bespoke campaigns that exploit misconfigurations and weak identity controls.

Meanwhile, the expansion of IoT and the rollout of 5G technologies are widening the attack surface. Smart factories, logistics systems, and digital healthcare services introduce operational technology (OT) into the threat matrix, where legacy devices and long upgrade cycles make patching and segmentation difficult. The country’s growth in cloud services and data centres increases both exposure and potential blast radii for incidents.

Technology response: AI, cloud‑native security and platform thinking

Defenders are moving beyond single‑tool approaches. The market is trending towards AI‑driven detection, cloud‑native controls, and identity‑centric security. Organisations are investing in platforms that ingest telemetry from the cloud, endpoints, and identity systems and apply analytics to detect anomalies that humans would otherwise miss.

Also Read: When security fails, trust breaks: Why cybersecurity is now a business priority

Managed security service providers (MSSPs) and outcome‑based models have surged in popularity as firms seek to close the talent gap quickly. MSSPs offer 24/7 monitoring and triage, but outsourcing comes with its own risks: opaque performance metrics, dependency on third‑party operational models, and potential systemic exposure if a provider is compromised. Boards must avoid treating MSSP engagement as a checkbox; robust contract governance and independent validation are essential.

The role of local vendors and integrators

Thailand’s market is a hybrid of global vendor platforms and local implementers. Multinational firms bring research, scale and broad telemetry, but localisation, regulatory alignment and on‑the‑ground integration are frequently delivered by domestic system integrators and service firms. Local companies such as BullVPN, UpperVPN, and NotVPN are carving niches in VPN and enterprise security services, providing context‑aware solutions tailored to Thai enterprises.

This division of labour is pragmatic: global technology solves for capability gaps; local players ensure technology fits the market and regulatory context. Yet it also underscores a skills arbitrage — advanced threat research and high‑end engineering often remain concentrated in overseas teams, leaving Thai organisations dependent on imported expertise for complex incident response.

Talent: The invisible bottleneck

The recurring refrain across industry and government is simple: talent shortage. There is a dearth of cloud security engineers, threat hunters, red‑teamers, and forensic investigators. Educational institutions produce graduates, but not at the scale or specialisation required to staff continuous security operations. This gap manifests as longer detection times, inconsistent patching regimes, and heavy reliance on MSSPs.

Fixing this requires more than curriculum tweaks. Apprenticeships, industry placements, public‑private training initiatives, and retention incentives are needed to create career paths in cybersecurity. Without a pipeline that rewards advanced specialisation and keeps talent local, Thailand will continue to import critical skills at a cost to resilience.

Regulation, collaboration and the limits of compliance

Regulatory enforcement is intensifying, with compliance becoming a baseline requirement for participation in certain sectors. Public‑private cooperation has evolved from advisory forums into more operational partnerships, but information sharing remains inconsistent. For compliance to translate into real security, it must be bound to operational maturity: incident simulations, shared threat intelligence, and sectoral playbooks.

Compliance alone will not deter sophisticated attackers. As Tracxn summed it up succinctly: “Weaknesses persist in talent and SME readiness.” Regulatory frameworks can raise the floor, but the ceiling is determined by investments in people, process and cross‑domain platforms.

Industrial risk: OT, 5G and the cost of fragmentation

Industrial digitalisation highlights the cost of fragmented security stacks. Securing hybrid environments that span cloud apps, mobile endpoints and industrial controllers requires unified visibility and consistent policy enforcement. Legacy OT devices with limited security capabilities complicate segmentation and incident response. The need for an integrated platform security — rather than a patchwork of point tools — is urgent for organisations that rely on connected operations.

What happens next

Thailand’s cybersecurity market is maturing but remains fragile. The coming years will be pivotal: success depends on translating regulatory momentum into operational muscle. That means building local talent, standardising cross‑sector information sharing, investing in cloud‑native defences powered by AI and resisting the temptation to treat MSSPs as a panacea.

Also Read: Asia’s new cyber threat: AI that speaks your language

A sober reality check: progress is meaningful but reversible. Without sustained investment in people, processes and unified platforms, rapid digital adoption risks amplifying systemic vulnerability rather than resilience. The country must do more than adopt world‑class technology; it must embed world‑class execution.

“Weaknesses persist in talent and SME readiness,” Tracxn observes, crystallising the central dilemma. Thailand can design frameworks and buy technology, but execution — the grunt work of detection, iteration and accountability — will determine whether growth becomes a liability or a source of durable strength.

The post Thailand’s cybersecurity boom has a weak core appeared first on e27.

Groundup AI, the Singapore-born startup that pioneered what it describes as the world’s first Agentic AI for Cognitive Maintenance, has closed its most significant commercial agreement to date. The contract, valued at more than US$10 million, involves the large-scale deployment of the company’s flagship platform across multi-site critical operations for one of the world’s leading asset-heavy organisations.

The identity of the client has not been disclosed for strategic reasons, but the scope of the agreement signals growing institutional confidence in AI-driven approaches to industrial uptime and asset management. The deal was secured following a rigorous competitive evaluation process.

Traditional predictive maintenance has long been the industry standard, offering organisations advanced warning of potential equipment failure. Groundup AI’s proposition moves a step further, positioning its platform as a system capable of autonomous diagnosis, reasoning, and operational guidance. The startup terms it Cognitive Maintenance.

The platform is underpinned by the Groundup AI Asset Library, a proprietary repository of more than 5,000 anomaly signatures that enables deep-tier pattern matching and root cause analysis. The company’s in-house AI, GINA AI, continuously learns from live operational environments, improving diagnostic accuracy over time without requiring lengthy onboarding or complex deployment cycles.

“The world is ready for Cognitive Maintenance that reasons, diagnoses, and guides,” according to Leon Lim, CEO and Founder, Groundup AI, in a press statement.

Also Read: Human imposter syndrome magnified: When AI knows more than we ever could

Regional ambition

The record-breaking contract follows a period of rapid growth for Groundup AI. In April 2025, the company closed a US$4.25 million Series A funding round led by Tin Men Capital, with participation from Wavemaker Partners, SEEDS Capital, and HIVEN, the venture capital arm of CJ International Asia. That capital injection has allowed the company to accelerate its technical development and expand its footprint across the region.

Alex Wong, COO and Co-Founder of Groundup AI, described the contract as an opportunity to execute the company’s core mission at unprecedented scale. The firm has trained its sights on the manufacturing, maritime, and critical infrastructure sectors, which it believes hold billions in untapped value that smarter maintenance systems could unlock.

Groundup AI’s technology combines industrial Internet of Things infrastructure with proprietary acoustic sensors and autonomous AI agents to deliver what it calls Physical AI reliability. The platform is designed to bridge the gap between the domain expertise of experienced engineers and the pattern-recognition capabilities of machine intelligence.

Yong Tai, Sales Engineer and Founding Team Member, highlighted that the company’s ambitions extend beyond software. “From heavy industries to advanced manufacturing, this milestone is about finally bridging the gap between raw data and action on the ground,” he said.

As global industries face mounting pressure to reduce unplanned downtime, cut operational costs, and meet sustainability targets, the case for Cognitive Maintenance is becoming increasingly difficult to ignore. Groundup AI’s latest contract suggests that at least one major organisation is prepared to make that transition at scale — and that the era of autonomous industrial reliability may already be underway.

—

Image Credit: Groundup AI

The post How Groundup AI redefines what smart factories expect from their machines appeared first on e27.

Deeptech startups face a unique management challenge: how do you build a business around a technology that can take years to develop, when market trends and customer needs may shift every few months?

Big companies like Nvidia, Google, or Meta can take on these kinds of problems with large, highly specialised engineering teams and deep reserves of capital. Even then, success isn’t guaranteed. Startups don’t have that luxury. They operate with limited resources, and any sudden change in the market can render years of painstaking work irrelevant.

Conventional startup wisdom suggests analysing the market, identifying a niche, and then building a product to fit it. But in deeptech, this model often breaks down. By the time the technology is ready, the “perfect niche” may have disappeared or evolved into something entirely different.

Why invest in an engineering challenge over a market niche

When GPU Audio launched in 2017, the team made a deliberate choice: rather than chasing an existing niche, they decided to solve a fundamental engineering problem that could survive multiple shifts in the market. Their focus was on adapting graphics processing units (GPUs), which were originally designed for parallel rendering of graphics, to process audio data in real time.

It wasn’t a quick win. Between 2012 and 2015, the team built multiple prototypes and failed more often than they succeeded. A turning point came in 2017, when a leading researcher — a former advisor to Qualcomm’s founders and a recognised authority in ray-tracing engine design — joined the project.

Even then, it took years before the first working demonstrations appeared in 2020 and 2021, nearly a decade after early experiments began. During that time, GPU architectures changed, AI went mainstream, and consumer trends shifted dramatically. But the team stayed focused on their core technical goal, resisting the temptation to pivot toward short-lived opportunities.

In this case, prioritising long-term technological resilience helped the team navigate several market cycles and uncover opportunities that weren’t on the radar at the start. For teams in similar situations, it shows that this path — while risky and demanding a long planning horizon — can sometimes open unexpected doors.

The depth and complexity of deeptech engineering challenges

The obstacles GPU Audio faced illustrate why deeptech startups must often think differently. Two problems in particular stood out.

First, GPUs were never designed for audio. For graphics, small delays are acceptable — a dropped frame may go unnoticed. But audio requires near-instant precision. Even a tiny delay of just a few milliseconds can be audible, disrupting the experience entirely.

Second, the way sound is processed is fundamentally different from graphics. GPUs excel at handling millions of identical, independent tasks in parallel — the computational equivalent of a factory full of workers all performing the same action simultaneously. Audio, on the other hand, is a chain of small, interdependent steps. Each calculation depends on the result of the one before it. Getting GPUs to handle this kind of workload required more than optimisation — it demanded a reinvention of how audio processing itself could be structured.

Also Read: From pilot to scale: Why traditional VC metrics don’t work for climate deeptech

Challenges of this scale affect more than just the engineering roadmap — they shape how a team operates. Long development cycles call for clear, ongoing communication with both investors and internal teams. In our case, we made a point of sharing intermediate milestones to keep everyone aligned, even when the road to a finished product stretched over years.

Creativity in engineering: Borrowing ideas from adjacent fields

Without the deep pockets of a tech giant, GPU Audio needed more than persistence; they needed creativity. The team had to rethink audio algorithms from scratch, searching for ways to break down sequential tasks into parallelisable ones.

The breakthrough came by borrowing ideas from another domain: ray tracing in 3D graphics. The company’s chief scientist had extensive experience in this field, having built one of the fastest ray-tracing engines in the world. Ray tracing calculates reflections, shadows, and interactions across countless objects at once — problems not unlike the hundreds of processes required in real-time audio.

Applying these principles, the team built a new kind of audio process manager — a scheduling system that could aggregate audio streams, distribute workloads efficiently, and maintain the responsiveness required for real-time sound. What the industry had long dismissed as technically impossible suddenly became feasible.

Shifting from consumer products to developer SDKs

Solving the core technical problem was only half the battle. Next came the question of product-market fit. Breakthrough engineering doesn’t automatically translate into customer demand — especially in markets where preferences shift quickly.

Initially, GPU Audio released consumer software for musicians and sound engineers, tools that ran on GPUs rather than CPUs. While useful, this approach wasn’t scalable. The team realised that instead of building end-user products themselves, they could multiply their reach by offering a software development kit (SDK) to other application developers.

Also Read: Funding deeptech: Balancing potential and complexity in the search for capital

This shift made the technology more flexible, less tied to short-term consumer trends, and far more attractive to potential partners. It also created a pathway into industries that the founders hadn’t originally targeted. For some deeptech startups, shifting from direct-to-consumer products to an SDK model can provide more flexibility and make it easier to keep pace with changing industry needs — that was the case here.

Discovering unobvious opportunities

Moving to an SDK unlocked a surprising new vertical: the automotive industry.

Car audio systems were lagging behind broader automotive innovation. Even as electric vehicles, advanced infotainment systems, and autonomous driving became more sophisticated, most in-car audio processing still relied on outdated DSP chips. GPU Audio saw an opportunity to modernise this layer.

The company developed zoned audio technology — allowing different passengers to hear different content simultaneously without interference. A driver could take a call over the front speakers while children in the back seat enjoyed a movie, all without overlapping sound.

This innovation didn’t just improve in-car entertainment; it opened the door to entirely new use cases, from personalised multimedia systems to interactive voice-based services. It also showed how a deeptech startup could scale by partnering with established industries, repurposing its core technology to meet needs far beyond the original vision.

The takeaway here is to stay open to markets beyond the original target segment. Often, meaningful scale comes from industries that haven’t yet gone through a full wave of innovation.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. Share your opinion by submitting an article, video, podcast, or infographic.

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

Image courtesy: Canva Pro

The post Deeptech’s secret: Ignore the market, master the engineering, and let opportunity find you appeared first on e27.

In the startup tech world, inclusivity is a complex, often moving target. 

On paper, progress is remarkable: women-owned businesses are growing. In Singapore, for instance, the number of female-founded businesses has more than doubled over the past 15 years. Inside our office, I’m proud to lead an all-female startups and partnerships team dedicated to fuelling the next generation of innovators. We are proof of what happens when women are given the reins to build.

Yet, step outside our doors into the broader ecosystem, and the reality remains stubbornly stodgy. 

The “90 per cent problem”

We attend and run hundreds of events a year to connect founders with capital and resources. Despite our best efforts at outreach, general event attendance consistently skews 90 per cent male. Even within Aspire’s own thriving FoundersXchange community, participation remains disproportionately male.

The problem isn’t an appetite for innovation, because we’ve seen firsthand that when the environment changes, the faces change too. At a recently hosted Women in AI event, we welcomed more than 80 women—and the energy was absolutely electric. Seeing that many women in one room, dissecting the future of LLMs and neural networks, was deeply inspiring.

It proved that the talent is there, it’s just often sidelined by the default startup culture. It also led me to wonder why we don’t see this reflected in broader tech spaces? Why does a “Women in AI” tag draw a female crowd, while a general “AI Summit” doesn’t? The answer isn’t that women aren’t interested. It’s that the general tech ecosystem has become coded with a specific brand of bravado—and when we host a women-focused event, the “imposter syndrome” that many feel in a room that is 90 per cent male disappears.

Also Read: Bridging the gender gap in GenAI learning: Strategies to get more women involved

Tackling the gap

More broadly, the lack of women in the room translates directly to a lack of capital in the bank. Early-stage funding for female entrepreneurs is in free fall. 

According to the latest Tracxn data, the drop-off is staggering. Singapore’s early-stage funding for female entrepreneurs fell by 39 per cent. Across Southeast Asia, total funding for women-led startups plunged from US$871.8 million in 2022 to just US$198 million in 2024 — a far steeper decline than that seen among male-founded businesses. Perhaps most alarming: in 2024, no late-stage deals were recorded for women-led startups in Singapore.

The irony here is proving costly. Women are objectively more capital-efficient than their male peers. Boston Consulting Group research shows that for every dollar of funding, women-founded startups generate 78 cents in revenue – more than double the 31 cents generated by male-founded firms. 

Beyond making the funding environment more equitable, we are learning a few things ourselves in our quest to create a more diverse community:

• Women often feel they have to over-index on expertise just to speak up in male-dominated rooms. Specialised events, like the Women in AI session, provide a psychological safety that general events seem to lack. 
• Networking has an inherent algorithmic bias. If the majority of startup and tech communities are male, their referrals will likely be male. We have to be hyper-intentional about breaking these closed loops.
• Finally, diversity in visibility, attendance and panels shouldn’t be a “nice to have” check box – it should be a key metric. It is a lead indicator for capital. When women aren’t seen at important summits, they aren’t seen by investors. 

We cannot afford to treat female participation as a once-a-year celebration in March. To the men who make up the 90 per cent in our rooms: It is time to question the absence. Don’t just “support” inclusivity; demand it. If you’re invited to a panel that is entirely male, ask who is missing. If your network is a closed loop of the same voices, break it. 

—

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

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

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

The post The 90% blind spot: Bridging the tech ecosystem gap appeared first on e27.

Think about your neighbourhood convenience store. Its secret sauce is simple: it’s everywhere. By saturating every street corner and office cluster, these stores become a natural part of our daily rhythm, optimising logistics while staying top-of-mind for consumers.

But petrol station convenience stores (C-stores) have always played by a different set of rules. They didn’t pick their spots because of retail foot traffic; they are there because, well, that’s where the fuel tanks are. For decades, we only stepped inside because we had to fill up. It was convenient by necessity, not by choice.

But the winds are shifting. And they’re blowing far beyond the smell of petrol.

The great pivot: From “pump and go” to “stay and consume”

As the global energy transition picks up pace, the traditional “big oil” retail model is facing a structural shake-up.

McKinsey expects the global fossil-fuel retail value pool to shrink from US$88 billion in 2019 to US$79 billion by 2030. Meanwhile, Non-Fuel Retail (NFR) is the new frontier, with Asia expected to be one of the fast growing regions.

What’s driving this is straightforward:

• EV adoption is extending dwell time. The customer journey is shifting from a 3–5 minute “pump-and-go” stop to a 20-minute-plus “plug-in-and-wait” visit.
• Fuel margins are under pressure. Price transparency and intense competition keep spreads thin, leaving little room for profit growth at the pump.

Together, these changes flip the logic of the site. When customers stay longer, the shop stops being a “nice add-on” and becomes central to protecting site-level profitability.

Global benchmarks: The “retail-first” future

In mature markets, the transition is already well underway. Across Europe and North Asia (Japan/South Korea), over 80 per cent of stations are integrated retail hubs, with non-fuel sales contributing roughly 40 per cent of total profits.

The US model is perhaps the most evolved. With over 90 per cent of stations operating C-stores, the “Forecourt + QSR (Quick Service Restaurant) + Coffee” combo is the real breadwinner. The pumps outside are essentially a loss-leader to drive traffic to the high-margin fresh food and coffee inside.

Also Read: Singapore’s green future – Are homes and condominiums ready for EVs?

The China context: Scale meets opportunity

China presents a different picture: very large networks, but non-fuel productivity is still catching up.

By footprint, Sinopec’s Easy Joy is reported at around 28,000 stores, and PetroChina’s uSmile at over 20,000. But scale alone does not guarantee retail strength.

Take Sinopec as an example. The company rolled out its non-fuel strategy in 2006, partnered with McDonald’s on drive-thru concepts that same year, established its Easy Joy convenience chain in 2008, and entered the freshly brewed coffee segment in 2019. In its 2023 annual report, Sinopec disclosed that non-fuel revenue reached tens of billions of RMB, just 2.3 per cent of the marketing and distribution segment’s total revenue, yet contributed roughly 18 per cent of the segment’s profit (more than RMB 4.5 billion (US$625 million)).

This proves that the future isn’t in the tank. It’s on the shelves.

The way forward: Evolving into lifestyle “super-nodes”

The EV revolution is doing more than just changing engines; it is creating a brand-new “Dwell Economy.” When a 3-minute fuel stop turns into a 20-to-40-minute charge, the “ceiling” for what you can sell a customer rises exponentially.

To capture this, the forecourt must evolve. Here are the four shifts that matter most:

From “one store” to “multiple missions”

Forecourts are adding missions quickly: coffee and hot food, car wash and basic automotive services, parcel drop-off and pick-up, and even small community services. Operating models are evolving too, with more shop-in-shop, leasing, and partnerships.

A simple way to frame it: forecourts are learning to monetise time, not just transactions. Every extra minute a customer spends on-site is a fresh opportunity for margin.

A natural fit for instant retail

Petrol stations have built-in advantages for on-demand fulfilment: high visibility, easy roadside access, ample parking, and existing retail space. For any brand building a last-mile network, the forecourt is the ultimate distribution node.

This is why we’re seeing major delivery platforms partner with fuel retailers like Easy Joy. By turning the forecourt into a local fulfilment point, the station becomes a vital part of the city’s logistics layer, serving customers who might not even be on the premises.

Also Read: Is India on the verge of shifting gears to EVs?

Digital is no longer optional

In forecourt retail, digital used to mean faster checkout and more accurate inventory. That is now table stakes. The bigger value is improving decision-making at scale.

To navigate this, Sinopec’s Easy Joy partnered with Dmall, a global provider of AI-driven retail digitalisation solutions, to move beyond basic automation to true data-driven execution.

This isn’t just about “smart shelves”. It’s about building a “People-Vehicle-Life” ecosystem. By unifying data across fuelling, charging, and dining, the station stops waiting for random purchases and starts predicting them.

Think of it as a dynamic intervention: the moment a driver initiates a charge, the system calculates their 30-minute dwell time and pushes a tailored offer, perhaps a fresh meal combo or a car grooming service, at exactly the right moment. This digital infrastructure transforms a passive wait into a high-margin, highly operational consumption window.

Conclusion

Forecourt retail is increasingly a combined play across energy, convenience retail, and last-mile fulfilment.

In the future, the best sites won’t be defined only by how much fuel they sell. They’ll be defined by how well they convert traffic into baskets, how many missions they can serve, and how effectively they use the customer’s time on site.

Fuel will still matter, but it won’t be the whole story.

—

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 post Beyond the pump: Why the future of fuel retail isn’t about fuel appeared first on e27.

We are used to identity being a human problem. A person signs in, gets assigned roles, and systems enforce access based on policy. Even when we talk about “non-human identities,” the mental model still tends to be infrastructure: service accounts, API keys, workload identities.

Agent-to-agent interaction breaks that model.

In the emerging architecture of AI-integrated platforms, agents will not only assist with one product. They will interact with external agents, negotiate APIs, coordinate tasks across tools, and execute actions that span organisations. 

This is barely discussed today, which is exactly why it deserves attention.

Why is this different from traditional integration

Cross-platform integrations are not new. What changes is the nature of decision-making.

Classic integrations are deterministic. A webhook fires. An API is called. A workflow runs. The system does what it was programmed to do.

Agents introduce delegation and interpretation. They decide what to call, when to call it, and how to combine outcomes. They reason over ambiguous inputs and incomplete context. They also learn patterns from interactions over time. That means “correct behaviour” is not just a matter of validating a token. It becomes a matter of validating intent, scope, and safety in motion.

When an external agent calls your agent, you are not just receiving a request. You are accepting an upstream decision.

The core identity question: Who is the actor?

With humans, the actor is clear. With service accounts, the actor is a system you control. With agents, the actor becomes layered.

Is the actor the user who initiated the request? The agent who interpreted the request? The platform that hosts the agent? The organisation that deployed it? Or the chain of agents that influenced the final action?

In real systems, it will often be all of the above. Without a shared way to represent that chain, we will end up with brittle trust based on convenience: “This request came from a reputable provider, so it must be fine.”

That is not a security model. It is a hope model.

Also Read: Embracing AI evolution: The crucial role of data management and cybersecurity in AI success

We need delegation integrity

Authentication tells you who is calling. It does not tell you whether the caller has the right to ask for what they are asking.

Agent-to-agent systems will need to prove not just identity, but delegation. The receiving system should be able to answer:

• Who delegated this action?
• What was the approved scope?
• What constraints were in place?
• What context was used to make the decision?
• How recent is the authorisation, and can it be revoked?

Today, most inter-org trust collapses into static secrets, broad OAuth scopes, and contractual assumptions. Those mechanisms were designed for services, not for autonomous decision engines.

Authorisation becomes dynamic and contextual

In a multi-agent world, authorisation cannot be a single static gate. It has to be context-sensitive and risk-aware.

If an external agent is asking to pull a file, the risk depends on the file type, its sensitivity, the destination, the current anomaly signals, and the actor chain. If an external agent is asking to trigger a workflow, the risk depends on blast radius, downstream integrations, and reversibility.

This forces a new discipline: designing “agent actions” as a controlled interface, rather than letting agents operate through broad administrative permissions. If your agent can do anything your user can do, you have effectively created a second user with fewer human constraints.

The trust boundary will shift from “app” to “action”

The safest mental model is that identity moves from being account-centric to action-centric.

Instead of granting an agent broad access to a system, you grant it the ability to perform specific actions under specific constraints. Each action has a policy. Each action is logged with intent and provenance. Each action can be throttled, sandboxed, or reversed.

This is already how high-trust systems are built. The difference is that it will need to become mainstream, because agents will otherwise accumulate privilege faster than governance can keep up.

Decision cascades in multi-agent systems

Agent-to-agent trust is only half the challenge. The other half is what happens when agents form chains.

Future systems will call other agents and trigger downstream automations. 

The failure mode here is not “one wrong answer.” It is “one wrong answer that becomes an input signal for ten other systems.”

Also Read: The new cybersecurity battlefield: Protecting trust in the age of AI agents

Cascades are not hypothetical

Organisations already have cascading automation. A monitoring alert triggers a ticket, which triggers an on-call action, which triggers a deployment rollback. The difference is that these chains are built from deterministic rules.

Agents make the chain probabilistic.

If an agent misclassifies an event, it may call the wrong downstream tool. If it overconfidently infers intent, it may trigger a workflow that was never meant to run. If it misreads context, it can propagate that error through multiple dependent actions.

The scary part is that each step in the chain can look locally reasonable. The system “followed the process.” The process was simply driven by a flawed inference.

Why we lack containment models

Traditional containment models assume discrete incidents: isolate the host, rotate credentials, block the IP, patch the vulnerability.

Cascades do not behave like that. They are distributed and asynchronous. They cross product boundaries. They may involve third-party agents. By the time you notice something is wrong, the downstream effects have already occurred in multiple systems.

This is why we need cascade containment models. Not as an abstract research area, but as an engineering requirement for systems that allow agents to trigger actions.

Principles for cascade containment

A mature cascade model starts with acknowledging that not every agent output should be executable.

Some practical principles are worth adopting early.

• Bounded autonomy: Agents should have clear limits on what they can execute without confirmation. Those limits should tighten as the blast radius grows.
• Progressive trust: An agent earns autonomy through reliable behaviour and predictable outcomes over time, not through initial configuration. New agents, new integrations, and new workflows should start constrained.
• Circuit breakers: If an agent triggers unusual rates of actions, unusual cross-system combinations, or repeated failures, automation should pause. This is deliberate friction that appears when the system deviates from normal.
• Risk scoring at the edge: Each action request should be evaluated not only by identity, but by context and potential impact. High-impact actions should require stronger proof and stricter gating.
• Explicit rollback paths: Actions that are hard to reverse should require higher certainty. If rollback is easy, you can allow more autonomy.
• Provenance and traceability: Every agent decision that leads to an action should carry a trace of what triggered it, what context was used, what downstream calls were made, and what constraints were applied. Without traceability, containment becomes impossible.

Users will demand autonomy, then blame it

As agents become more capable, the pressure to let them act will grow. Users will want “just handle it” experiences. And when something goes wrong, the same users will be surprised that the system acted without permission in a nuanced case.

This is why guardrails cannot be an afterthought. They have to be part of the product contract. The system should clearly communicate what it can do autonomously, what it will ask before doing, and how it will behave under uncertainty.

The goal is not to reduce automation. The goal is to make autonomy safe.

—

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 post Agent-to-agent trust: The next identity challenge appeared first on e27.

In startups, promotions rarely begin during the review cycle.

They begin much earlier, in the moment a manager decides who gets the hard project, who joins the important meeting, and who is trusted to represent the team when the stakes are high.

Most companies do not describe it that way.

They say they promote on merit. They say performance speaks for itself. They say the best people rise. But in many startups, what gets called merit is often built through access first.

Culture decides who gets seen

Workplace culture is often treated like a soft issue.

It sits next to values decks, off-sites, and internal messaging. It is discussed as something important, but not always as something operational.

That is a mistake.

In startups, culture acts more like infrastructure. It shapes who gets visibility, whose judgment gets tested, and who becomes familiar to leadership before promotion discussions even begin.

This matters because formal review processes usually come late. By the time someone is being discussed for a bigger role, the room is often not deciding from scratch. It is confirming a view that has been forming for months.

By then, the bet has usually already been made.

The hidden feeder system into leadership

Most startups have a hidden feeder system into leadership.

It is not written down. It does not show up in the org chart. But it is there.

It looks like stretch assignments. It looks like cross-functional launches. It looks like being invited into difficult conversations. It looks like being trusted with work that carries risk and visibility.

The people who get these opportunities build a record that later looks like leadership readiness.

The people who do not get them may still perform well. In many cases, they are the ones keeping the team stable. But they are often seen as reliable executors rather than future leaders.

That is where culture starts shaping outcomes.

The real question is not only who performs well. It is who gets the kind of work that later gets rewarded.

Also Read: From lead generation to pipeline hygiene: What startups often miss

How this plays out inside startups

This pattern does not always look dramatic.

That is part of the problem.

Picture a lean startup preparing for an important product launch. It needs someone to coordinate across product, marketing, operations, and leadership. The project is messy. It is visible. It comes with pressure.

A manager picks someone they already feel comfortable with. Maybe that person is available late into the night. Maybe they are already part of informal leadership circles. Maybe they simply communicate in a style that feels familiar in the room.

Another team member, equally capable, stays on steady execution work. They keep things moving. They deliver. They are dependable.

A few months later, one person is described as showing leadership potential.

The other is described as strong, but not quite ready.

Nothing openly unfair may have happened in that moment. No policy had to be broken. No one had to say anything discriminatory.

But the outcome is still not neutral.

The people who get the biggest opportunities early are usually the ones the company later calls naturally ready.

Who benefits, and who gets left behind

This system tends to reward people who are already closer to power.

That can mean people with easier access to senior leaders. It can mean people who are more comfortable speaking in high-status settings. It can mean those who can mirror the pace, style, or availability patterns of the people already in charge.

It can also quietly disadvantage people with caregiving responsibilities, people who are newer to influential networks, and people whose strengths show up more in depth than self-promotion.

Women and other overlooked talent often feel this gap without always being able to name it.

They are told to speak up more, be more visible, or act more strategically. But what they often need is not better advice. They need fairer access to the work that creates visibility in the first place.

That is why culture cannot be reduced to tone or sentiment.

Culture decides who gets the proving ground.

Why this becomes a growth problem

Some leaders still treat this as a fairness issue alone.

It is a fairness issue. But it is also a growth issue.

When people see that advancement depends more on informal access than clear opportunity, they stop trusting the system. Some disengage. Others leave. The strongest often leave quietly, after they realise the ceiling is lower than the company admits.

That weakens retention.

It also narrows the leadership pipeline. Companies keep selecting from the same profiles, then wonder why their bench feels thin.

The cost shows up elsewhere, too. Teams lose morale when effort and opportunity drift apart. Hiring gets harder when internal stories about advancement start circulating outside the company. Product and growth decisions become narrower when the same lived experiences keep getting rewarded and elevated.

Once the same people keep getting the proving ground, the company starts calling the result merit.

That is when a cultural problem becomes self-defending.

Also Read: Data-driven or gut-led? Why the best startups do both

One structural shift that matters

Startups do not need perfect systems to improve this.

But they do need to stop treating opportunity as invisible.

One useful shift is to make stretch assignments visible and review who gets them.

That does not mean removing manager judgment from every decision. It means paying attention to patterns that are usually left unexamined.

• Who is getting cross-functional projects?
• Who is getting exposure to senior leadership?
• Who is repeatedly trusted with strategic work?
• Who is staying essential but unseen?

When leaders track opportunity flow, they can spot whether growth is being distributed fairly or simply recycled through familiarity.

That is a better place to start than waiting for performance review season and trying to correct an outcome that was shaped months earlier.

If a company wants fairer promotion outcomes, it has to look at the path to readiness, not just the final score.

What founders should sit with

Most startups do not set out to build uneven leadership pipelines.

They do it more quietly. They confuse familiarity with readiness. They confuse availability with commitment. They confuse informal trust with objective judgment.

Then, over time, careers, pay, and authority begin stacking on top of those early choices. At that point, the system feels harder to question because so much already depends on defending it.

That is why workplace culture matters more than companies often admit.

It does not just shape how people feel at work. It shapes who gets built into the future of the business.

—

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 post Why merit in startups is often decided before performance reviews begin appeared first on e27.

The image of a street food seller in Bangkok or a vegetable vendor in Mumbai with a worn-out QR code on a cart is what people think of when they talk about “New Asia”. For the people who created the economy, this is a big deal. It shows that the future is here.

However, when we look closer, we must pose a question: Does the vendor’s experience on the issue of digital payments relate to the experience of a large retailer? When the countries of India and Southeast Asia transition to a cashless future, people tend to confuse access and fairness. It is not enough to provide a merchant with a wallet or a QR code.

A good system ensures that the technology behind the wallet is not only operational for a small business owner but also functional for a large company. The digital economy is no longer fair. They are the ones who have the lowest margins, who put their money into having the greatest risks.

The “silent failure” problem

Speed is the most important thing in a world where payment is made on time. It is unbelievable that India contributes to 46 per cent of the total real-time transactions in the world. However, to a trader, payment speed is not the greatest issue.

Also Read: Value creation: The US$3T private equity blind spot

It is because they are not sure. This uncertainty occurs when the account of a customer is debited. The merchant does not receive notification. According to a recent study conducted by a fintech company, the percentage of this issue is approximately 1.8 per cent of digital transactions. And worst of all, when a few transactions are in process at a time.

A failure rate of 1.8  per cent is a pain for the accounting department of a company. To a small business owner who has missed a payment, it does not mean that they cannot afford to buy food or to stock up on the day. These vendors cannot wait 48 hours before the money is refunded. Failure by the system comes at a cost to the merchant. They are forced to inform the customer that his or her money is lost. They are also unable to deliver goods to them.

Constructing the safety net

Addressing this imbalance requires looking beyond user adoption and examining the architecture that powers digital payments. A transaction is rarely a direct exchange between two individuals. It typically moves through a network of banks, gateways, switches, and settlement systems before it is completed.

When disruptions occur within this chain, large enterprises often have dedicated teams, technical redundancies, and financial buffers to manage the impact. Smaller merchants usually operate without these safeguards. For them, even brief uncertainty around payment confirmation or settlement can affect working capital, inventory decisions, and customer trust.

Also Read: How product design is democratising access to growth‑stage equity

This has led to growing interest in infrastructure-level approaches that improve transaction reliability and transparency. One such approach is payment orchestration, which focuses on coordinating multiple payment pathways and service providers within a unified operational framework. Rather than relying on a single processing route, orchestration layers can help detect network issues, reroute transactions when needed, and provide clearer visibility into transaction status.

The significance of such systems lies less in technological sophistication and more in their potential to reduce operational friction for merchants. When payment flows become more predictable, businesses can spend less time resolving failed transactions and more time focusing on growth and service delivery.

Improved resilience also helps protect margins. The costs associated with lost sales, delayed refunds, and manual reconciliation can be disproportionately high for small enterprises. Strengthening the reliability of payment infrastructure, therefore, becomes not just a technical priority but an economic one.

As digital adoption deepens across emerging markets, equity will depend not only on expanding access to wallets and QR codes but also on ensuring that the systems behind them function consistently for participants of all sizes. A cashless ecosystem becomes more inclusive when confidence in transaction completion is shared by both informal vendors and large retailers.

—

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 post Beyond the QR code: Why infrastructure is the real key to digital equity appeared first on e27.