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.