Artificial Intelligence (AI) has become the cornerstone of digital transformation, enabling businesses to automate tasks, enhance decision-making, and drive innovation. At the heart of this revolution lies Large Language Models (LLMs)—powerful AI systems capable of understanding and generating human-like text.

While LLMs offer tremendous potential, organisations face a critical decision: should they train an LLM from scratch, or fine-tune an existing model and integrate Retrieval-Augmented Generation (RAG) for enhanced performance?

Training an LLM from scratch demands extensive resources, including massive datasets, powerful computing infrastructure, and deep AI expertise. On the other hand, fine-tuned LLMs combined with RAG offer a more cost-effective, scalable, and efficient alternative that allows businesses to harness AI’s power without the heavy upfront investment.

As organisations navigate the complexities of leveraging artificial intelligence, they face a critical dilemma: What is the most effective approach to unlock its full potential and drive maximum impact — build from scratch or optimise?

The rise of domain-specific AI: Why one size doesn’t fit all

AI has evolved significantly from being a broad, general-purpose tool to an increasingly specialised and industry-focused solution. While massive, general-purpose models like GPT-4, Gemini, and Llama are versatile, they often lack the domain-specific expertise that businesses need.

For example, an AI model trained for financial fraud detection requires specialised knowledge of banking regulations, risk assessment, and transaction patterns—something a generic LLM may struggle to provide.

This has led to the growing adoption of fine-tuned AI models, where organisations take a pre-trained LLM and customise it with industry-specific data and knowledge. By fine-tuning a model, businesses can significantly improve accuracy, relevance, and efficiency while ensuring AI aligns with their unique requirements.

Additionally, Retrieval-Augmented Generation (RAG) is emerging as a game-changer, allowing LLMs to fetch real-time data from external sources, ensuring the model’s responses are not limited to static training data.

The true cost of training an LLM from scratch

For companies considering training an LLM from the ground up, the costs and challenges are staggering. Training a state-of-the-art language model requires massive computational power, vast datasets, and a dedicated team of AI researchers and engineers. Estimates suggest that models like GPT-4 cost over US$100 million to develop, requiring thousands of high-performance GPUs and TPUs running for months.

Also Read: Re-skilling in the age of AI and navigating the future of work in Malaysia

Beyond the financial burden, training a model from scratch also demands access to high-quality, diverse datasets. Without extensive, well-curated data, organisations risk producing biased or inaccurate AI models. Even after training, maintaining and updating the model is an ongoing challenge—requiring periodic retraining to keep the AI relevant in fast-changing industries like finance, law, and healthcare.

For most enterprises, the reality is clear: training an LLM from scratch is impractical, costly, and unnecessary. Instead, a more efficient approach is to leverage pre-existing models and customise them through fine-tuning and RAG.

Fine-tuned LLMs + RAG: The smarter, scalable alternative

Rather than reinventing the wheel, organisations can take advantage of pre-trained LLMs and fine-tune them with industry-specific data to align with their business needs. Fine-tuning significantly reduces the time, cost, and computational power required compared to training from scratch, while still delivering a model that is highly accurate and domain-specific.

What makes fine-tuning even more powerful is its synergy with Retrieval-Augmented Generation (RAG). Traditional LLMs rely solely on their pre-trained knowledge, which becomes outdated over time. However, RAG enables AI to retrieve real-time data from external sources, knowledge bases, APIs, or proprietary datasets—ensuring that responses remain accurate and relevant. This is particularly valuable for industries like financial markets, healthcare, and cybersecurity where up-to-date information is critical.

By combining fine-tuning with RAG, organisations can deploy AI that is cost-effective, faster to implement, continuously updated, and contextually aware—without the burden of retraining a model from scratch.

Challenges in implementing LLMs and RAG

While fine-tuning and RAG present a more feasible AI strategy, organisations must navigate key challenges to ensure success. One of the biggest hurdles is data quality and availability. Fine-tuning an LLM requires high-quality, well-labeled datasets, but many organisations lack structured and curated data sources. Inadequate data can lead to biased, inaccurate, or unreliable AI models.

Integrating RAG also adds complexity, as it requires organisations to establish real-time data retrieval pipelines. This often involves using vector databases, which require specialised expertise to set up and maintain. Additionally, businesses must address security and compliance concerns, ensuring that sensitive proprietary data remains protected and does not get exposed through external retrieval mechanisms.

Despite these challenges, fine-tuned LLMs and RAG remain a superior option, especially when organisations invest in robust data strategies, secure deployment practices, and continuous model monitoring.

The business case: Why fine-tuned LLMs + RAG outperform training from scratch

For most organisations, the decision to fine-tune an LLM and implement RAG is a no-brainer. The advantages extend beyond just cost savings—businesses also gain greater flexibility, faster deployment times, improved accuracy, and enhanced real-time capabilities.

Financially, fine-tuning and RAG drastically cut down AI development costs. Instead of spending millions on compute resources, organisations can fine-tune pre-trained models for a fraction of the cost, often using cloud-based AI services like AWS Bedrock, Azure OpenAI, or Hugging Face’s model hub. Deployment is also exponentially faster, with fine-tuned models ready for production in weeks instead of years.

Moreover, the real-time adaptability of RAG gives businesses a major competitive edge. A financial institution can use RAG-powered AI to pull the latest stock market trends, while a cybersecurity team can retrieve real-time threat intelligence. This ability to update AI knowledge dynamically without retraining the entire model makes fine-tuning with RAG the ideal approach for organisations that operate in fast-changing environments.

Also Read: The human touch endures: Why AI won’t replace all blue-collar jobs

How organisations can get started with fine-tuned LLMs + RAG

To maximise AI’s potential, businesses should start by identifying key areas where AI can deliver the most value—whether it’s automating customer service, enhancing research, or improving decision-making. Next, organisations should select the right pre-trained LLM, such as OpenAI’s GPT models, Meta’s Llama, or open-source alternatives like Mistral or DeepSeek.

Once a model is chosen, the fine-tuning process begins. Businesses should gather high-quality, domain-specific data and use supervised fine-tuning to align the model’s responses with their industry requirements. Following fine-tuning, RAG can be integrated to ensure AI has real-time access to external knowledge sources.

Leveraging an integrated, pre-configured AI technology stack like QXLI can significantly reduce costs, minimise the need for specialised resources, and accelerate time-to-market. By streamlining deployment, optimising infrastructure, and automating key processes, organisations can focus on innovation rather than technical complexities, ensuring faster AI adoption and improved operational efficiency.

For deployment, organisations must also consider security, compliance, and scalability. AI solutions can be hosted on-premises or in the cloud, depending on data privacy needs. Finally, continuous monitoring and optimisation should be implemented to ensure AI models remain accurate, fair, and up-to-date over time.

The clear winner in the AI game

While training an LLM from scratch may seem like the ultimate AI achievement, the reality is that it is costly, time-consuming, and impractical for most organisations. Instead, the smarter and more efficient path lies in fine-tuning pre-trained models and integrating RAG, enabling businesses to build highly specialised, real-time AI solutions with minimal cost and effort.

By adopting fine-tuned LLMs with RAG, organisations can achieve unparalleled AI performance, scalability, and adaptability—without the burden of massive infrastructure investments. In the race for AI dominance, organisations that adopt efficient, agile, and real-time AI strategies will not only stay ahead of the curve but also define the future of their industry.

—

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.

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

Image credit: Canva Pro

The post AI without the price tag: How fine-tuned LLMs + RAG give you more for less appeared first on e27.

Revenue-per-employee just became the only metric that matters—everything else is theatre.

“Scale the signal, not the size. The graveyard is full of startups that confused growth with bloat.”

Here’s what keeps venture partners up at 3 AM: Goldman Sachs needs 49,000 bodies to generate what Apple does with surgical precision per worker. So why do our sharpest startups morph from lean execution machines into bureaucratic nightmares the second Series A money hits the bank?

Check your cap table, then check your payroll. Which one still has ammunition for the next pivot? If that answer takes longer than five seconds, you’ve already loaded the gun pointed at your own runway.

The numbers don’t lie, they just kill

Apple cranks out US$2.38 million per employee. Meta delivers US$2.19 million. Nvidia hits US$2.06 million. These aren’t vanity numbers—they’re proof that peak performance demands precision, not padding.

Instagram sold to Facebook for US$1 billion with just 13 employees—that’s US$77 million per person. WhatsApp’s 55-person team served 900 million users when Facebook acquired them for US$19 billion. Meanwhile, your Series-B peers are burning US$2M supporting 212 people to serve maybe 275,000 users.

Death by committee

CB Insights autopsied 966 startup corpses from 2020-2024 and found the same cause of death: premature scaling beat out pricing screw-ups, customer churn, and regulatory disasters. TechCrunch reported 966 shutdowns in 2024 versus 769 in 2023, and the pattern is clear. The casualties aren’t companies that never found product-market fit—they’re the ones that found it, then hired themselves to death.

The math is brutal. A US$2M runway keeps five elite performers fed for 24 months. Same cash with 20 average hires? 6 months, maybe less. Revenue per head crashes from US$400K to US$100K overnight.

Also Read: AI integration field notes for tech startups and scale-ups: Software engineering, product, and beyond

The physics of bureaucracy

Meta-analysis of 3,200 firms shows every new hire cuts total output by 0.9 per cent, with complexity growing exponentially—not linearly. This isn’t poor management—it’s physics.

UCLA tested this with LEGO assembly: two-person teams finished in 36 minutes while four-person teams needed 56 minutes. Panasonic’s own factory data shows productivity falls off a cliff after 50 workers per line.

Forbes surveyed 1,842 founders and found the decision paralysis that kills startups :

92 days is a full product cycle. While 100-person startups debate features, 10-person competitors ship, test, and iterate.

The institutional amnesia problem

Professor Jennifer Mueller at UC San Diego calls it “relational loss”—the support individuals feel as teams balloon. But there’s something deeper happening: institutional amnesia.

Winners forget they’re supposed to disrupt incumbents, not become them. Dropbox scaled from 4 employees to IPO by maintaining extreme focus on core product rather than building departments. Zoom’s Eric Yuan kept the company lean through US$100M ARR by prioritising engineering excellence over organisational complexity.

The consultant hack

McKinsey studied autonomous teams and found something counterintuitive: external expertise crushes internal mediocrity :

Triple the quality, 40 per cent of the cost, 2.6x faster. Smart startups get this. Instead of building marketing departments, they hire world-class agencies for campaigns. Instead of legal teams, they partner with top firms. They buy results, not résumés.

Stripe famously used this approach through Series B, keeping headcount under 50 while processing billions in payments by leveraging best-in-class partners for compliance, fraud prevention, and regional expansion.

The disruption paradox

If IBM’s 300,000 employees, Google’s 190,000, and Microsoft’s 220,000 could out-innovate small teams, venture capital wouldn’t exist. These giants would own every breakthrough, every disruption, every advance. Startups would be irrelevant.

But startups win through speed, precision, and agility. You don’t beat a 1,000-piece orchestra by building a 1,001-piece orchestra. You beat it with three snipers who know exactly where to aim.

Gallup found that 42 per cent of employees at sub-10 companies report engagement versus 30 per cent at larger firms. The Ringelmann Effect proves individual output drops as group size grows—researchers call it “social loafing”.

The vanity metric trap

Startups obsess over headcount as a growth signal, revealing a fundamental misunderstanding of value creation. ScienceDirect research confirms companies optimising revenue per employee significantly outperform headcount optimisers.

PitchBook analysed 1,100 tech exits and found revenue-per-employee optimisers outperform headcount growers by 240 per cent in exit multiples.

Consider the contrast: Snapchat reached 100 million daily users with under 100 employees, while traditional media companies needed thousands to serve smaller audiences. Signal serves 50 million users with fewer than 50 employees, proving that encrypted messaging at scale doesn’t require enterprise headcount.

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

The US$2M choice

Every seed extension forces a decision:

Each extra hire is equity that can’t go toward product, pricing, or distribution. Reid Hoffman warned Stanford students in 2024: “Hiring wrong at the wrong time is the fastest death sentence in venture. There’s no second product-market fit”.

The survival framework

The next decade’s winners won’t scale fastest—they’ll scale smartest :

• Hire top one per cent talent, not average performers
• Track revenue per employee, ignore headcount
• Buy expertise externally, not loyalty internally
• Build systems before adding bodies
• Resist the “we look bigger” vanity play

The mirror question

Tape this to your monitor: “If Goldman needs 49,000 people to match Apple’s margins, why are we copying Goldman?”

When that note falls off, so does your runway.

Every successful startup faces the same choice: keep the lean culture that created initial wins, or join the 90 per cent that scale into oblivion. Small teams already work—WhatsApp, Instagram, and Stripe proved that.

The real question is whether founders have the nerve to ignore conventional wisdom long enough to build something that lasts.

—

The post The scaling paradox: Why elite startups abandon their winning formula appeared first on e27.

When I first started my career in the early 2000s, I was given what felt like a painfully mundane task: enter postcodes into an internal business website, extract data on telephone exchanges, and type it into a spreadsheet to design metropolitan networks.

Being the bright-eyed graduate I was, I decided there had to be a better way. Over one weekend, I wrote a small Visual Basic app that took postcode inputs from the spreadsheet, screen-scraped results from the website, and parsed the data back. What once took two weeks could now be done in five minutes. That small hack freed me to spend more time with customers, and it helped win millions of dollars in deals. It also earned me a stern lecture from the IT team, who saw my tool as a denial-of-service risk.

That was my first taste of automation. Eliminating repetitive work did not just save time. It shifted where I could add value.

The rise of no-code automation

Fast forward to today, and we have tools like Zapier and n8n. Instead of writing code, anyone can drag and drop “if this, then that” rules to automate workflows. APIs that once needed developers are now accessible through simple interfaces. Internal processes that used to take hours can be chained together in minutes.

Once you have tasted automation, the next question comes quickly: Can we do more? Can we make these flows less brittle, more adaptive, even self-healing when something changes? Can we move beyond scripts toward systems that understand goals and figure out the steps themselves?

That is where I see agentic AI entering the conversation. It feels like my childhood dream of a Star Trek computer coming to life.

What agentic AI really means

Agentic AI refers to tools that do not just follow rules but act autonomously to achieve objectives. The leap is from executing tasks to making decisions. At its core, agentic AI is:

• Goal-oriented: An agent decides what actions to take to achieve an outcome. Example: “Reduce overcrowding by 15 percent this year” might lead it to adjust escalator directions, modify turnstile access, or recommend alternative exits.
• Adaptive and flexible: If one method fails, it explores alternatives instead of stopping with an error code.
• Reasoning-capable: It weighs trade-offs, infers missing information, and runs simulations on historical data to propose better steps.
• Autonomous: It runs continuously and can initiate workflows without explicit human triggers. A train delay, for example, could prompt an agent to adjust signage and entry gates to prevent bottlenecks.
• Multi-tool orchestration: It selects and connects tools across systems, even ones you did not hard-code to work together.

If automation is about rules, agentic AI is about goals.

Also Read: The digital lag: How traditional consulting is failing to grasp the agentic AI revolution

The data bottleneck

For all the excitement, one truth remains: an agent is only as good as the data it receives.

Take smart cities. Train stations are busy, complex hubs that are prone to overcrowding and often need manual intervention when something goes wrong. CCTV is everywhere, but cameras have blind spots, struggle in low light, and cannot reliably capture depth. The result is incomplete, sometimes misleading data.

At Curium we deployed LiDAR sensors across European stations to provide accurate insights into footfall, crowding, and hazards. An investor once asked me, “So what? Is this just a nice to have?” The question stuck with me. Not everyone sees why LiDAR adds value, or how agentic systems depend on high-fidelity baseline data.

When you add LiDAR to CCTV, the picture changes. You get full coverage, depth perception, and real-time accuracy. That coherent data unlocks predictive analysis: what is normal, what is not, and how conditions are evolving. With agentic AI, you can then test interventions dynamically, shift escalator directions, close entry gates, reroute passengers, and watch outcomes in real time.

What changes in practice

Forward-thinking operations managers are starting to see the opportunity. With richer data, agentic AI moves from buzzword to practical decision support. It is not only about efficiency. It is also about safety and resilience.

Imagine:

• Highway systems that adjust traffic lights and lane allocations dynamically.
• Public transport hubs that respond to surges before they spill into dangerous overcrowding.
• Driverless systems that adapt in real time to disruptions.

These are actual early signals of how work, infrastructure, and customer experience will be reshaped.

What leaders should pay attention to

The implications go beyond technology. Teams will shift from doing tasks to overseeing autonomous systems. Culture will need to balance trust in agents with accountability when mistakes happen. Regulators will ask the hard question: if an AI agent makes the wrong call, who is responsible?

Practical steps to take now:

• Start small: Pilot in low-risk domains with clear success criteria.
• Design for oversight: Keep a human in the loop, define escalation rules, and log decisions for review.
• Harden the data layer: Invest in sensor quality, data validation, and observability before you attempt autonomy.
• Focus on customer trust: Efficiency gains mean little if people do not feel safe relying on the system.

We are already seeing first steps in Singapore with fixed-route autonomous shuttles in Punggol. The lesson is simple: the path to autonomy starts with reliable data and measured rollouts.

Also Read: Agentic AI in action: How Southeast Asia’s startups are turning constraints into strengths

Trust starts from the baseline

I often show a simple pyramid when pitching autonomous systems. At the base is sensor data, above it sits automation and agentic AI for perception and decision making, and at the top is public trust.

If the base is weak, the whole pyramid crumbles. Faulty data leads to flawed decisions, which lead to failures that erode trust. Get the basics right: high-fidelity data acquisition and validation, and the rest of the pyramid can stand.

Public perception ultimately rests on outcomes: did the car crash, was the station overcrowded? By building robust data foundations before handing more authority to AI agents, we stand a better chance of not only achieving success but also earning lasting trust in the systems we build.

—

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 From automation to Agentic AI: Trust starts at the data layer appeared first on e27.

The extended downturn in Southeast Asia’s tech investment market has played out unevenly across different deal stages, revealing a distorted picture of market health due to the increased reliance on private debt and convertible notes.

While investment volumes have plummeted, median valuations and deal sizes surprisingly held up across stages well into the second half of 2024.

Also Read: SEA funding wiped out: Back to 2016 levels after historic slump

Per a Cento Ventures report, this perceived stability, however, was less a sign of market strength and more a distortion of the underlying correction. The rising use of private debt and convertible notes, often not reflected in standard public datasets, kept the number of priced equity rounds unusually low. This financial cushioning mechanism has obscured the true pace of the correction, despite a noticeable spike in startup closures and founders scrambling to achieve profitability.

The mismatch is apparent: headlines often hint at an “over-correction,” yet the valuations on record still appear sticky. The market is not yet clearing because founders have been reluctant to reset their valuation expectations, while investors are equally hesitant to meet those expectations. Debt instruments have provided a temporary bridge across this gap.

The core stack collapses

The capital cycle demonstrated sharp differences in pace across the venture stack:

• Series C+ deals slowed sharply as early as the second half of 2022, primarily due to the contraction in the supply of mega-deal capital. This gradual slowdown in Series B and C continued throughout 2022–2024, influenced by SEA-centric funds slowing down their deployment.
• Seed and pre-Series A stages initially saw a surge, buoyed by later-stage funds moving downstream in search of value, but this activity slowed dramatically by the first half of 2024.
• Series A and B stages proved the most resilient, holding up the longest, and only beginning to show a steep decline towards the second half of 2024.

The most concerning development arrived in mid-2024, marking a full reset of core investment activity back to 2016 levels. The previously stable seed to early Series B part of the core stack collapsed abruptly to one-third of its volume recorded during the 2021-2022 boom.

The subsequent decline was sharpest in the seed and pre-Series A segment (deals ranging from US$0.5 million to US$3 million), which reached a new low for the region after peaking in the first half of 2023. Similarly, the Series A to early B segment (deals greater than US$3 million up to US$10 million) was down by 50 per cent year-on-year.

Valuation and deal size nuances

Despite the drastic reduction in volume, median valuations did not experience a uniform collapse, thanks largely to the low number of priced rounds.

For pre-Series A deals, the median pre-money valuation held steady at US$5 million in 2024, showing zero per cent change compared to 2023. However, the median deal size for pre-SeriesA doubled in 2024 compared to 2023, reaching US$1 million, even as deal volume dwindled. This suggests that only larger, highly selective pre-A rounds were being completed.

Series A median valuation reached US$20 million in 2024, an increase of 5 per cent over 2023. This slight uptick towards the end of H2 2024 was partially attributed to increased liquidity made available by the Malaysian government’s fund-of-funds initiatives. Median Series A deal size remained stable at US$4.0 million between 2023 and 2024.

Series B median valuation saw a significant adjustment in 2022, dropping from US$90 million to US$55 million. In 2024, however, Series B median valuation increased by 17 per cent year-on-year, reaching US$52.0 million. This rise was primarily driven by pricier deals in the health-tech and Digital Financial Services sectors during the second half of 2024.

In contrast to the valuation rise, the Series B median deal size continued its gradual slide, falling to US$8 million in 2024, reflecting how capital from later-stage funds no longer “spilled over” into earlier stages.

The relative stability in valuations indicates that the supply of and demand for early-stage venture deals have declined at similar rates following the major adjustment of 2022.

Shifting country valuations

A look at country-specific Series A median valuations shows a convergence towards the US$20 million to US$25 million range. By 2024, the valuation gap between the lowest median (Thailand) and the highest (the Philippines) was 2.9X. Indonesia’s Series A valuations remained relatively stable despite the overall market slowdown, aided by increased bridge financing and venture debt that helped companies avoid downrounds. Malaysia stood out as one of only two markets with rising Series A valuations in 2024.

Also Read: IPO surge, unicorn scarcity: The new face of SEA’s funding landscape

Series B valuations also showed significant growth in several key markets for 2024. The Philippines saw a massive 85 per cent increase in Series B median valuation, while Indonesia saw a 25 per cent rise, and Malaysia witnessed a huge 102 per cent increase in 2024 compared to 2023.

This suggests that only the strongest companies, or “market survivors”, secured priced Series B rounds in 2024, creating an upward shift in market averages across these regions. However, the continuous use of venture debt and bridge financing rounds caused a continuous drop in Series B deal volume, reducing the clarity of the valuation signal.

The post When debt replaces equity: How SEA startups mask a funding winter appeared first on e27.

When it comes to structuring smart, capital-efficient deals, it’s hard not to notice how Singapore Special Purpose Vehicle (SPV) setups are gaining traction across Asia. Investors love the stability, clarity, and jurisdictional strength Singapore offers.

From startup founders to the sophisticated LPs, the buzz is real: Singapore is staking its claim as Asia’s SPV powerhouse. According to ACRA, there were 459,263 companies registered in Singapore as of August 2025, a net increase of more than 23,000 over the past 12 months. This steady rise underscores the growing adoption of Singapore Private Limited Companies (Pte Ltds), the go-to structure for SPVs.

Clearly, the numbers are pointing to Singapore’s steady momentum as a hub for efficient, investor-friendly structures. So why are syndicate leads, venture investors, and family offices choosing Singapore over traditional offshore centres? Let’s break it down.

Singapore SPV structures explained

An SPV (Special Purpose Vehicle) is a separate legal entity created for a specific project or investment. Think of it as a clean container: it holds assets, pools capital, and isolates risk.

For investors, this structure unlocks:

• Risk mitigation: Assets and liabilities sit within the SPV, not the parent entity.
• Flexibility: SPVs can back one deal or manage a portfolio.
• Tax efficiency: Optimised through Singapore’s favourable regime and treaties.
• Global participation: Easier access for investors across borders.

Singapore SPVs are already widely used in venture capital (to pool capital into single deals), private equity (to hold international assets), and cross-border deals (to leverage tax treaties and legal infrastructure).

Also Read: Singapore Fashion Council backs Loom Carbon to tackle textile waste

Why investors choose Singapore

Singapore isn’t just another offshore option for investors; it’s a full-fledged global financial hub.

• Strategic location

Positioned at the heart of ASEAN, Singapore connects Asia to global markets. Its role as a regional hub makes it a strategic base for cross-border investment flows.

• Stable governance

Singapore’s political stability, efficient regulatory framework, and pro-business policies provide the predictability and security investors value in long-term commitments.

• Financial infrastructure

The country’s ecosystem includes top-tier banks, experienced legal and tax advisors, and respected regulators. Together, these elements create a reliable foundation for structuring complex transactions.

• Speed of setup

Establishing a Singapore SPV is fast and efficient. In most cases, investors can set up an entity and secure a functional bank account within 1–2 days, a significant advantage for time-sensitive deals.

• Global acceptance

Singapore SPVs are widely recognised and accepted globally. With the exception of jurisdictions sanctioned by MAS, investors from nearly any country can participate, and these vehicles can deploy capital across international markets with ease.

Tax and regulatory advantages of Singapore SPVs

One of the biggest draws of a Singapore SPV is its tax framework.

• Double Tax Treaties (DTAs): Singapore has signed treaties with 90+ countries, including India, China, the US, and the UK. This means lower withholding taxes and protection against double taxation.
• Corporate tax rates: A competitive flat rate of 17%, with partial exemptions for startups and SMEs.
• No capital gains tax: Profits from selling shares, property, or financial instruments are often exempt.
• Tax incentives: Programs like the Global Trader Programme encourage cross-border activity.

Compared with the Cayman Islands (increasingly criticised for opacity) and BVI (perceived as unstable under new regulatory pressures), Singapore stands out. Investors see it as both efficient and credible.

Also Read: Singapore tops global AI hiring charts: One in six jobs now reference AI

Singapore vs Cayman vs BVI

Feature	Singapore SPV	Cayman Islands	BVI
Tax transparency	Strong, OECD-compliant	Under scrutiny	Moderate, facing pressure
Reputation	High legitimacy	Tax haven label	Less stable
Regulatory complexity	Predictable, supportive	Rising compliance	Increasingly complex
Double tax treaties	90+ treaties	Few	Limited
Capital gains tax	None	None	None

Efficiency and cost benefits of setting up an SPV in Singapore

Setting up a Singapore SPV is also about making operations smoother.

Efficiency benefits:

• Centralised investor management: Pool multiple investors into one entity for easier communication and reporting.
• Streamlined transactions: Create a dedicated vehicle for M&A, asset transfers, or single-deal investments.
• Operational freedom: Lighter regulatory burden than a parent company, allowing flexibility.
• Evolution of digital banking: The rise of digital banks has further accelerated SPV efficiency. Licensed by MAS, providers like Finmo now open SPV accounts in a matter of hours, fully compliant with regulatory standards.

Cost benefits:

• Reduced taxes: DTAs cut withholding taxes, boosting net returns.
• Lower admin costs: By outsourcing SPV management to specialised providers, investors cut overhead.
• Optimised financing: SPVs can raise debt or issue securities at better terms than their parent entities.
• Risk containment: Liability is isolated within the SPV, protecting the wider portfolio.
• No capital gains tax: Singapore imposes no capital gains tax, allowing investors to retain a larger share of their returns.

Singapore wins for investors seeking both efficiency and legitimacy. According to Singapore’s Ministry of Finance, the country’s Exchange of Information on Request (EOIR) regime has been rated as “Compliant,” the highest possible designation under the OECD’s international tax transparency standards.

Final thoughts

For syndicate leads and investors, the message is clear: Singapore is the future-proof home for your SPVs. From favourable tax treaties to unmatched credibility, setting up your SPVs in Singapore combines efficiency with trust, qualities global investors now prioritise.

Compared with traditional offshore centres, Singapore provides the infrastructure, reputation, and legal strength to structure complex deals at scale. If you’re building syndicates, raising cross-border capital, or managing specialized projects, Singapore is where the smart money is setting up.

—

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: DALL-E

The post Why Singapore SPVs are becoming Asia’s go to structure for smart investors appeared first on e27.

What if your startup had a team member who never sleeps, works 24/7, makes smart decisions, learns from experience, and needs zero breaks or benefits? Now imagine having not just one, but a fleet of these team members.

This isn’t science fiction. It’s now a real option for startups thanks to a new evolution in artificial intelligence: agentic AI.

Agentic AI systems go far beyond the traditional AI that passively waits for user input. These next-gen digital agents operate independently. They set objectives, make decisions, plan multi-step tasks, execute them fully, adapt to shifting inputs, learn from outcomes, and keep improving over time. They emulate the actions of skilled human team members, yet perform at speeds, precision, and consistency that human teams simply can’t, especially under conditions of resource scarcity.

For early-stage startups, this is game-changing.

Startups are often strapped for time, limited in team size, and operating without the budgets or bandwidth of mature firms. The challenge is constant: how do you cover more ground without overwhelming people, raising burn rates, or compromising quality? You need to move fast, execute flawlessly, and adapt continually — all while hiring slowly or not at all.

This is where agentic AI steps in — not just as a tool, but as a strategic multiplier.

Let’s take a closer look at how agentic AI is transforming the way startups scale across their most critical functions:

Customer support that grows with you

Customer questions, tickets, and issues don’t scale slowly — they pop up fast as you acquire users. Agentic AI helps you stay ahead. Intelligent agents can autonomously manage thousands of support tickets, resolve repetitive issues, and refine their responses as they learn from interactions.

They don’t just follow scripts — they adapt over time. By handling common questions with speed and accuracy, AI enables human agents to focus on complex or sensitive requests, providing a better customer experience while avoiding bottlenecks.

AI-driven sales and marketing funnels

Startups don’t just need leads — they need the right leads, engaged and nurtured without draining bandwidth. Agentic AI can screen inbound leads, score them, personalise outreach, track engagement, and launch follow-up campaigns — all without human supervision.

These agents essentially act as SDRs and marketing ops bundled into software. Their ability to craft tailored messaging based on user behavior means higher conversion rates with fewer human hours involved. For scrappy teams trying to punch above their weight in crowded markets, this unleashes big-league capabilities.

Also Read: The digital lag: How traditional consulting is failing to grasp the agentic AI revolution

Operational systems that don’t need babysitting

Startups operate in a constant state of flux. Meetings move. Team members wear multiple hats. Processes evolve week to week. Agentic AI helps make operations smoother – even when things get chaotic.

Whether it’s scheduling interviews, managing internal workflows, syncing calendars, or coordinating vendor logistics — the AI observes, learns, and adapts in real-time. It becomes a backend brain optimising task handling, reallocating resources, and eliminating bottlenecks before they hit your inbox.

Resilient supply chain and logistics, without extra ops hires

For product-based startups managing inventory, fulfilment, or delivery operations, supply chain hiccups can become deal-breakers. Agentic AI now enables real-time inventory forecasting, automated restocking approvals, and smart logistics routing that reacts on the fly.

Your digital agents don’t just trigger alerts — they resolve issues, suggest alternatives, and even negotiate with vendors when programmed to do so. The result? You can run more like a scaled ops team without needing to build one.

Why now? The tools are finally accessible

This shift is powered by advances in large language models (LLMs), prompt engineering, and easy-to-integrate no-code tools. You don’t need a PhD in machine learning or a large engineering team to get started. Tools like GPT-4, n8n, Make, or Zapier can now be wired together into performance-ready intelligent agents.

What used to take quarters, budgets, and full developer teams can now be achieved in weeks — sometimes days. We’re at a moment where skilful design and integration matter more than scale of headcount.

The strategic advantage: Scale without bloat

Hiring can’t always keep up with growth, and budgets don’t always stretch to hire every expert. Agentic AI gives founders a new scaling strategy — one where every process becomes smarter and faster, without multiplying team size. When implemented right, AI doesn’t just ‘save time’ — it allows early-stage startups to act like companies twice their size.

It means your five-person company can accomplish the output of 20. Your late-night backlog? Handled by agents overnight. Your onboarding process? Automatically personalised, delivered, and tracked. Your metrics dashboard? Refreshed daily — by the AI itself.

Also Read: From hype to harmony: Why agentic AI needs a platform-first mind-set to redefine CX

This isn’t about replacing people. It’s about up-levelling your entire team.

Each agent becomes a support system, enabling humans to spend more time on uniquely creative, strategic, and customer-focused work. Where once you needed a marketing specialist, analyst, and admin, now a small founding team can cover all these roles intelligently — with help from digital agents.

From vision to execution

The biggest challenge for founders isn’t knowing AI exists. It’s knowing where to apply it first. That’s where clarity is key: map your most repetitive tasks, high-friction workflows, or time drains. Then apply the Promptive Method (Discover, Build, Optimise) to develop agents custom-fit to your startup’s exact needs. Smart implementation multiplies value and sets your company on a different growth trajectory — without bogging you down in technical complexity.

Startups that embrace agentic AI in 2024 aren’t just trimming costs — they’re accelerating growth. They’re turning workflows into competitive advantages, transforming onboarding into seamless automation, and turning data into action without delay.

Agentic AI is the co-founder you didn’t know you could hire: always on, always learning, never asking for equity. And it’s here to help you scale smarter.

—

The post From 5 to 50: How agentic AI lets startups operate like enterprises appeared first on e27.

Hiring marketing specialists is notoriously tough for early-stage startups, especially for founders without extensive business experience. How can they be sure the marketer they bring on board will deliver results? Will a performance marketer drain the budget on ineffective social channels? Could a content manager accidentally damage the company’s reputation? And do you really need to hire a big-name CMO from a large firm, or might external consultants be the smarter choice?

To mitigate these uncertainties, it’s crucial that founders first develop a clear marketing strategy. Only after understanding the company’s needs and deciding on the type of marketing team required should they begin the hiring process.

Finding the right specialist with industry-relevant experience and skills can be challenging and, for entrepreneurs who are new to this, prone to missteps. While there is no one-size-fits-all formula—since every startup’s marketing needs differ—here are four key pitfalls founders should avoid when recruiting marketers.

Don’t hire without defining clear marketing goals

It’s easy to feel compelled to quickly build a marketing team once you secure funding. I’ve seen startups hire multiple marketers at once, only to drastically cut back weeks later. For example, one direct-to-consumer startup hired 10 full-time marketing staff after their seed round, but just two months later, reduced the team to four—realizing that external partners could handle much of the workload efficiently.

My advice: take the time to evaluate exactly what marketing tasks need to be done and the skills required to accomplish them. Avoid rushing to hire a well-rounded marketer or several specialists without clarity. Instead, distinguish which projects could be outsourced, which roles require full-time hires versus freelancers, and match candidates’ experience to your business goals.

Set specific, measurable objectives for your marketing hires upfront. This clarity will help you identify the right candidates and objectively assess their performance over time.

Also Read: Balancing personalisation and privacy in business marketing

Don’t overlook team fit and company culture

Marketing professionals rarely work in isolation—they collaborate with developers, product managers, designers, and others. Especially in small startups, it’s vital to ensure new hires align with your company culture and can work well with existing team members.

Founders should be hands-on during the hiring process from the beginning, participating in interviews and assessing how a candidate might integrate with the team dynamics. Choosing someone who meshes well with your team can prevent costly misunderstandings and foster a more harmonious work environment.

Don’t rely solely on credentials and past titles

While formal qualifications and experience matter, they shouldn’t be the sole basis for your hiring decision. Early-stage startups thrive on shared values and growth mindset.

Sometimes, hiring a junior marketer with potential and cultural alignment is smarter than recruiting a seasoned executive whose perspective clashes with the team. You can always mitigate a less experienced hire’s risks by seeking advice from external experts.

This approach not only builds a cohesive team but also creates opportunities to mentor and develop talent internally.

Also Read: Mastering the craft: 5 essential tips for elevating your B2B marketing game

Don’t overcomplicate the hiring process

Many marketers say that hiring procedures have become unnecessarily drawn-out and stressful. Some startups implement numerous interview rounds and test assignments, which may deter strong candidates.

Ask yourself if all those steps are truly essential. Can you streamline the process—reducing interviews and making prompt hiring decisions when you find a good fit?

Delays in recruiting can mean missed opportunities; a capable marketer starting sooner can drive value and growth even as they continue to learn.

Early-stage startups usually need to move quickly to scale, so an overly complex recruitment process can work against this goal. Focus on efficiency by conducting only the necessary interviews, eliminating redundant steps, and aligning hiring with your startup’s priorities.

The fast pace and resource constraints of early-stage startups make marketing hires a crucial and challenging task. Founders must balance thoroughness with agility, aiming to build an effective marketing team that can accelerate growth without unnecessary risk or delay.

By setting clear goals, respecting team culture, valuing potential over credentials, and keeping the process straightforward, founders can significantly improve their chances of recruiting the right marketers to power their startup’s success.

The post 4 common hiring mistakes to avoid when building a marketing team for your early-stage startup appeared first on e27.

The narrative guiding Southeast Asia’s tech investment was profoundly revised during 2023-2024, according to a new Cento Ventures report. The narrative moved decisively away from broad, consumer-led growth models towards the dominant and resilient story of digital financial services (DFS).

Since Indonesia’s valuations peaked in early 2022, investors have cycled through various competing regional growth stories, most of which failed to materialise under closer scrutiny.

Also Read: SEA funding wiped out: Back to 2016 levels after historic slump

Two prominent narratives were tested and found wanting:

• Vietnam as the “next China”: This theory lost significant momentum as political shifts throughout 2023 and 2024 dampened investor confidence.
• The Philippines as the “next Indonesia”: This thesis came under scrutiny after closer analysis of Indonesian middle-class economics cast doubt on the long-term sustainability of consumer-led investment strategies, especially following failed Initial Public Offerings (IPOs) of large consumer companies.

These proposed growth stories relied heavily on assumptions of robust consumer spending, which ultimately proved less resilient than investors had initially anticipated.

The retreat from Indonesia and rise of the Philippines

The narrative shift has significantly impacted capital allocation across Southeast Asia’s key markets. Notably, Indonesia, long seen as the region’s scale driver, has received less than its “fair share” of investment in the regional digital economy since the second half of 2023. Funds previously explicitly raised for “consumer story” ventures in the archipelago are now shifting away from tech investments and moving towards mid-cap private equity-style investments, such as F&B chains.

In stark contrast, the Philippines has re-emerged as a key investment destination, propelled by the new dominant story of digital financial services. The country’s environment–characterised by a unique combination of light-touch regulation and significant financial disparities–has acted as a catalyst, accelerating its financial sector forward.

This shift has been evidenced by major capital injections into digital banking competitors, with companies such as Salmon, UNO Digital Bank, Mynt, and PayMaya all securing substantial funding in 2024.

While receding in overall VC investment share, Indonesia continues to provide critical scale for digital lenders that are experimenting with different operational models, both with and without established bank charters.

The super-app thesis abandoned

A major strategic pivot across leading digital platforms in 2024 confirmed the death of another growth narrative: the “super-app” model. Most digital platforms have officially abandoned the multi-vertical “super-app” thesis to concentrate on originating and distributing financial services.

The era of non-digital financial services super-apps was officially deemed over by 2022. This strategic refinement means that digital finance has become the key driver of profitability announcements across the region’s leading platforms throughout 2024. The concept of multi-vertical (or diversified services) accounted for US$866 million of capital invested in 2023–2024, representing 13 per cent of the top five sectors.

Business automation and exit resilience

Beyond the DFS dominance, another sector showing a burst of activity is business automation. While financial services captured 48 per cent of total capital invested in the top five sectors in 2023-2024 (US$3.3 billion), business automation accounted for US$507 million, or 7 per cent.

Also Read: When debt replaces equity: How SEA startups mask a funding winter

The spike in investment into business automation reflects a surge of experimentation with hybrid B2B marketplace plus SaaS business models across a variety of industries. However, the capital invested in business automation is back to its baseline after a brief B2B marketplace-driven spike.

Regarding exits, the ecosystem is still reeling from exposed instances of fraud and financial mismanagement across various companies. While the overall recovery remains a work in progress, more minor mergers and acquisitions (M&As) are still occurring. The market generally does not clear for significant acquisitions until founders and late-stage investors reach a point of desperation due to misaligned expectations.

A notable exception in early 2024 was the Tokopedia-ByteDance sale, which occurred under significant regulatory pressure. Despite these challenges, median exit valuations saw a substantial increase in 2024 compared to 2022, suggesting that strong, albeit smaller, exits are still being achieved.

The post Why Southeast Asia’s consumer-led growth story has officially ended appeared first on e27.

Many midlife professionals and creators have decades of ideas and experiences worth sharing. But turning those insights into content that markets their skills often feels overwhelming.

Writing feels stressful. Video production feels too technical. The result? Great ideas stay private, while social media feeds remain empty.

A real case: An artist and a dusty window

I was speaking with an artist friend in his late fifties. He described a striking moment: rain and dust forming patterns on an old window, like nature painting its own artwork. His explanation was vivid, but he dreaded writing it into an artist statement.

So I opened ChatGPT, tapped the microphone, and asked it to simply talk. Within minutes, AI transformed his spoken words into a polished artist statement, even producing a Chinese version.

Then I told him: This is not just your statement. This is your Facebook post. This is your Instagram caption. This is how you promote your art and your story online.

From idea to content pipeline

We pushed further. I asked ChatGPT to generate a short script for a video presentation of his concept. With CapCut, I combined the script and visuals. By the end of the afternoon, his inspiration had turned into:

• A professional artist statement
• A ready-to-publish social media post
• A short video for showcasing his work

This is the power of AI when you use it as a content pipeline. One idea, expressed once, becomes multiple marketing assets.

Don’t get stuck in fear

His first reaction was one many midlifers share: “If AI can do this, won’t it replace copywriters, marketers, or even me?”

This fear is driven by headlines and rumours, focusing on replacement instead of opportunity. The truth is simple:

• AI cannot create meaning without human input.
• It needs your ideas, your perspective, your story.
• What it offers is speed, scale, and structure.

Also Read: AI adoption evolves: Knowledge, confidence, and code creation soar

Instead of fearing replacement, midlifers should focus on innovation and leverage.

Why this matters for midlife professionals

In today’s market, visibility matters. Whether you’re an artist, consultant, or small business owner, social media is your storefront. AI gives you the ability to:

• Publish consistently without the stress of blank pages.
• Repurpose content across text, video, and multiple languages.
• Promote your skills skillfully by amplifying your unique voice.

For midlifers exploring a second act, this means you don’t need to wait for a marketing team or expensive campaigns. You can start today, with tools that are already available.

Final thought

AI won’t replace human creativity. But it is already replacing the excuses that keep ideas hidden.

Midlife professionals who open up, experiment, and use these tools will find themselves ahead: more visible, more agile, and more skillful in promoting what they do best.

The real risk is not AI itself — it’s refusing to explore what’s already possible.

Clarity in the Age of AI is about shifting from fear to leverage. Your next idea could already be tomorrow’s marketing campaign, if you let AI help you shape it.

—

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 of the author.

The post From idea to impact: How midlifers can use AI to turn inspiration into marketing content appeared first on e27.

The Singapore Institute of Technology (SIT) has inaugurated the SIT x NVIDIA AI Centre (SNAIC) at its Punggol Campus, positioning the facility as a central hub for innovation and talent development focused on applied artificial intelligence (AI).

The centre aims to foster cross-sector and cross-border collaboration through symbolic partnerships with major industry and academic players.

The SNAIC joint venture aims to accelerate AI adoption across the broader ecosystem via applied research collaboration. It seeks to nurture a robust pipeline of AI talent to support Singapore’s national AI strategy and its strategic goal of tripling the national AI workforce.

Also Read: Talents remain an issue in AI proliferation, but here are 6 steps that businesses can do to tackle it

The facility already hosts 19 industry doctorate and industry master’s (ID/IM) students pursuing studies through industry-applied research projects. These students work alongside full-time research engineers and experts from NVIDIA and SIT.

Strengthening Singapore’s talent pipeline

A key initiative launched alongside the centre is the SNAIC AI Programme, developed in collaboration with the Infocomm Media Development Authority (IMDA). This six-month programme, supported under the TechSkills Accelerator (TeSA) national initiative, aims to train over 200 AI practitioners, including fresh graduates and mid-career professionals, within the next three years.

The curriculum is split into two phases: two months of intensive AI training covering modules developed by SNAIC, including cutting-edge skills like Generative AI, Large Language Models (LLMs), Retrieval-Augmentation Generation, and Agentic AI. This is followed by four months of hands-on projects with industry partners, allowing participants to solve real-world business challenges using AI under expert supervision.

Kiren Kumar, Deputy Chief Executive of IMDA, said: “Through IMDA’s TeSA initiative, we are building Singapore’s AI talent pipeline by combining the best of industry expertise and academic excellence. With SIT-recognised certifications, and hands-on experience provided through leading company partnerships, our learners will graduate ready to tackle real-world AI challenges and contribute meaningfully to our digital economy.”

For those seeking to nurture advanced AI talent, SNAIC will also support the newly established Applied AI Doctoral Training Centre (AAIDTC). The AAIDTC aims to train 10 Industrial Doctorate students annually, focusing on complex, multi-year, industry-driven AI challenges that require developing novel methods beyond off-the-shelf tools. This system creates a tiered talent development pathway from practitioners to applied AI researchers and innovators.

Driving innovation across key sectors

SNAIC is leveraging strategic partnerships to build a robust pipeline of industry-relevant AI projects, positioning Singapore as a global centre for applied AI innovation.

Public transport innovation (SMRT): SMRT Corporation has deepened its collaboration with SIT through the SMRT-SIT Transport Living Lab. Together with SNAIC, SMRT is co-developing advanced AI solutions aimed at boosting operational effectiveness and enhancing the commuter experience.

Specific initiatives include:

• A GenAI-powered system for safety investigations designed to automate the incident reporting process, benefiting staff involved in reporting and investigations, thereby enhancing operational efficiency.
• Project AiDiSA (AI-Driven Intelligent System) which enhances automated case creation and sentiment analysis to boost productivity.
• An AI-driven lifestyle recommendation engine developed for Wink+, a mobile app offering access to MRT maps, real-time bus arrivals, and food deals accessible by public transport.

Financial services transformation (Prudential): The Prudential AI Lab is incubating AI-powered solutions to deliver a better customer experience and significant business impact. This partnership focuses on co-developing innovative AI solutions, utilising Singapore’s talent pool to support Prudential’s goal of becoming a digital-first insurer.

Also Read: Singapore aims to lead in AI — but where’s the talent?

Regional academic gateway: To foster cross-border innovation, SNAIC has established partnerships with international universities, including Monash University (Australia), Chulalongkorn University (Thailand), and Vietnam National University–University of Economics and Law (VNU-UEL). These collaborations will facilitate joint research projects addressing cross-border industry challenges and promote talent exchange through joint seminars and workshops, positioning SNAIC as a crucial gateway for regional applied AI collaboration in Southeast Asia.

The post Singapore Institute of Technology and NVIDIA launch AI centre to boost talent and innovation appeared first on e27.