SG-based firms stepped up tech acquisitions in rest of SEA in 2022
According to DealStreetAsia, such tech acquisitions witnessed a 166% surge from 12 in 2021 to 32 in 2022; About 62.5% of the targets were Indonesian firms, mainly from fintech and financial services sectors.

Prenetics announces US$200M JV for early cancer detection
The joint venture Insighta with Dennis Lo, an expert in non-invasive prenatal testing and liquid biopsy, aims to develop and commercialise a breakthrough early detection screening test for several types of cancer.

Singapore-based Apeiron bags US$37M in debt funding
Apeiron collects various types of food and agricultural waste – including used cooking oil, tallow, and palm oil mill effluent – and converts them into biodiesel.

UTU raises US$33M to help travellers get more from tax-free shopping
It has also acquired CardsPal, a Singapore-based fintech that offers deals and promotions nearby to users; Utu’s goal is to create innovation in the tax-free shopping sector, which allows tourists to reclaim VAT on their purchases.

Byju’s assures investors of FY 2022 earnings audit by September
The Indian edutech unicorn previously filed its financial statements for FY 2021 in September 2022; Last week, Byju’s auditor Deloitte resigned due to long delays in completing its FY 2022 financials.

Amazon plans to invest another US$15B in India by 2030
The e-commerce group has invested about US$11B in India to date; The vast majority of the new capital is likely earmarked for AWS expansion in India.

Ant Group expands Alipay+ to all 7-Eleven stores in Malaysia
The integration, which was facilitated by Razer Merchant Services, allows the 2,400 7-Eleven stores in Malaysia to accept AlipayHK, GCash, Kakao Pay, and TrueMoney as payment options.

GoTo-backed Electrum to manufacture 250K e-bikes annually
It is constructing a factory in the West Java province to manufacture two-wheel EVs in Indonesia; These collaborations involve efforts in battery packaging and production, EV adoption, and battery swapping.

Alt-protein firm Green Rebel readies to serve the Philippines, Vietnam
The meat consumption in those markets has been relatively higher than in the rest of the region; People are willing to shift to plant-based meats if the quality is on par and affordable.

‘Starting with a clear culture in mind is a vital for companies’
Finhay CEO Huy Nghiem says short-term financial stability is as important as long-term goals; If we cannot meet the former, we can’t meet the long-term goal either.

How Salmon promotes financial inclusion with AI banking in PH
In its first year since launching, Salmon has achieved significant milestones, including launching its first point-of-sale lending product.

Solarad.ai can forecast accurate energy generation for solar plants
Solarad.ai’s software continuously updates the forecasts to account for real-time weather conditions and fine-tune the accuracy of the predictions.

The future of gamification: Connecting brands with consumers through games
Gamification solutions will bear greater resemblance to mobile games in which consumers can earn rewards through a curated process.

UK implements stricter rules: Crypto airdrops and dree NFTs banned
The UK concluded a consultation on new rules for the crypto sector and proposed an authorisation regime overseen by the FCA.

—

Fundraising or preparing your startup for fundraising? Build your investor network, search from 400+ SEA investors on e27, and get connected or get insights regarding fundraising. Try e27 Pro for free today

Copyright: puvasit

The post Ecosystem Roundup: Singapore tops tech M&As in SEA in 2022, Apeiron bags US$37M in debt funding appeared first on e27.

Baskit, a startup that focuses on digitalising and growing distribution businesses in Indonesia’s supply chains, today announced that it had raised US$3.3 million in seed funding.

The funding round comes from Betatron Venture Group, Forge Ventures, Investible, 1982 Ventures, DS/X Orvel, Michael Sampoerna, and other prominent global and regional angels.

It followed a US$1.5 million pre-seed funding round that Baskit announced in March.

With the fresh funding, the company plans to accelerate expansion for proven revenue streams while doubling down on its technology stack and leveraging resources to strike numerous important contracts with brand principals and manufacturers.

“We aspire to build a platform that orchestrates all of the relevant players in the supply chain, generating economic gain and ultimately benefiting consumers. To do this, we have a razor-sharp focus on building superior tech that wins in functionality and usability for the SMEs we support,” Baskit CEO Yann Schuermans said in a press statement.

Also Read: Echelon: Breaking barriers in Southeast Asia’s supply chain

“With the raised funds, we will accelerate both our tech roadmap and city-by-city expansion, continuously embedding flexibility into our platform to cater to Indonesia’s vast and diverse landscape.”

How Baskit drives growth to a new height

Launched in November 2022, Baskit said that it had seen rapid growth since its inception, sustaining growth rates of more than 70 per cent month-on-month during the period.

In an email interview with e27, Schuermans explained the keys behind this achievement.

“Growth comes from a carefully executed go-to-market in regions that needed our solutions the most. We used local relationships and referrals to drive quick adoption and an efficient acquisition process, and have been able to effectively monetise our software, revenue-share initiatives (we take a cut if our customers grow), and signed lucrative contracts with lenders and manufacturers who want to access our wholesalers.”

He also shared the user acquisition strategy that Baskit is implementing: “Acquisition strategy is a very localised one; we play city by city and ensure that we build critical mass in each one of the areas we operate in. We’ve been very focused on West Java, particularly Greater Bandung Area, for now, and will be expanding to Greater Jakarta Area and Central Java in H2 2023.”

According to the company, the rapid expansion is reflective of a real need in the market for strengthening the operations of distributors and wholesalers.

These two players are now facing increasing competitive and fiscal pressures in the aftermath of COVID-19.

Also Read: Logistics platform for e-commerce merchants Jumppoint raises US$6.5M

This “new normal” involves high awareness but low access to tech, thinning margins and additional operational complexity from rampant inflation, and declining sales from softness in certain sectors. Baskit believes that the timing is conducive for them to support across these areas.

Schuermans said, “Collectively, trade and industry account for more than half of Indonesia’s GDP and are supported by more than 200,000 traditional distribution businesses that ensure that people can access a wide array of products, from biscuits to cement for construction.”

“While it is fashionable to disintermediate these businesses, it is not sustainable as they hold valuable infrastructure and relationships. We are deeply committed to supporting these intermediaries in their pursuits of powering their local communities and believe that doing so will yield tremendous economic benefits in the long run.”

This year, there are three things that Baskit wants to focus on: Expansion into the aforementioned areas; product improvements, particularly around its SaaS product and data product capabilities; and expansion of the number of lucrative contracts to build a solid foundation for profitability into 2024.

—

Image Credit: Baskit

The post Baskit raises US$3.3M in seed funding to empower Indonesia’s distribution infrastructure appeared first on e27.

The rapid growth and investment in generative AI have sparked a race among startups to leverage this technology and build AI-native applications. In Q1 of 2023 alone, US$1.7 billion was invested in 46 generative AI companies. However, in the midst of this excitement, it is crucial for these startups to consider the long-term viability and competitiveness of their products.

As we reflect on the transition from the introduction of horizontal cloud software in the early-2000s to the emergence of industry-specific cloud software in the late-2000s, it becomes clear that specialised startups hold immense value.

Building defensibility

We see a few key considerations to think deeply about as startups think about how to build defensibility vis-a-vis the models and potential incumbents coming in. To avoid getting commoditised, there are two critical questions: does the startup provide enough value on top of the model layer? And on what basis does a startup develop its moat?

One of the fundamental considerations for startups is to assess whether they offer enough value on top of the model layer to avoid being commoditised by it. While generative AI models provide a strong foundation, startups must focus on incorporating AI value through prompt engineering and fine-tuning specific to their use cases.

However, it is equally important to evaluate whether the value provided by the startup will remain significant as models continue to improve. Startups should strive to go beyond the capabilities of the underlying models and offer unique features, insights, or services that differentiate them from competitors.

Incorporating private data and customer context into the generative AI model is a powerful way for startups to improve the quality and relevance of their outputs. By leveraging proprietary data sources and understanding customer needs deeply, startups can enhance the accuracy, personalisation, and value of their AI solutions.

Also Read: Solarad.ai can forecast accurate energy generation for solar plants, battery storage

AI’s value will be most powerful when tightly focused, so startups should aim to accumulate proprietary data. This strategy enables startups to create a competitive advantage that is difficult for incumbents or generic AI models to replicate.

The age of generative AI

In the current landscape, generative AI applications are following a trajectory akin to early cloud companies, albeit with a crucial distinction. Unlike the cloud era, where startups like Salesforce, Workday, and ServiceNow emerged as ground-up pioneers, the underlying technology of generative AI now closely aligns with industry giants such as Microsoft, Google, and Meta.

Moreover, with the advent of APIs and open-source models, the adoption of generative AI technology has become markedly more accessible, fostering an environment conducive to both incumbents and startups. Consequently, as these major players dominate the broad horizontal applications, startups must recalibrate their focus toward specific domains with narrow contexts.

To maximise the value of generative AI, startups should consider the optimal insertion point for their product within existing workflows. This requires understanding the pain points, inefficiencies, or opportunities where AI can make the most significant impact. By going deep into specific workflows, startups can minimise disruptions while unlocking the full potential of AI.

In the age of generative AI, the stronger your context window, the wiser your model and product. Founders need to have a deliberate emphasis on the emerging technical capabilities of generative AI, accompanied by a relentless pursuit of identifying functions or vertical problems that stand to benefit from their unique insights.

We’re at an inflexion point. Startups must learn from past platform shifts and understand the importance of narrowing their focus and leveraging proprietary data to create defensible businesses.

By providing value beyond the model layer, incorporating private data, going beyond incumbents, and finding the right insertion point, startups can build specialised AI applications that deliver superior outcomes for customers. In the era of generative AI, specialised software holds the key to success.

—

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 our e27 Telegram group, FB community, or like the e27 Facebook page

Image credit: Canva Pro

The post Unleashing the power of specialised AI startups in the era of generative AI appeared first on e27.

In October 2020, Vitalik Buterin shared his thoughts about a ‘rollup-centric roadmap’ for Ethereum. Looking at projects emerging at the time — Optimism, Arbitrum, zkSync, etc.— he wrote: “The Ethereum ecosystem is likely to be all-in on rollups (plus some plasma and channels) as a scaling strategy for the near and mid-term future.”

Buterin’s prodigious vision has become a reality due to the synergistic relationship between Layer-2 Rollups and Eth2’s enhanced data availability. Rollups are the Holy Grail of Ethereum scalability, potentially solving the scalability trilemma — i.e., blockchains must compromise security or decentralization to achieve scalability. 

Two types of Rollups are currently very common: Optimistic and Zero-Knowledge (ZK). But while both methods contribute significantly to Ethereum’s scalability, they have considerable limitations. Therefore, emerging ‘Smart L2’ platforms like Metis innovate ‘Hybrid Rollups’ that combine security and scalability while fostering decentralization. 

Experts at Metis, such as Co-Founder and CTO Yuan Su, have found that combining the fast finality of ZK Rollups and the transcendental scalability of Optimistic Rollups enables developer-friendly and secure Ethereum Layer-2 solutions. This is much needed for Web3’s sustainable and long-term adoption. 

The case for L2 Rollups

While Bitcoin lays the foundation for an alternative, censorship-resistant, peer-to-peer money, Ethereum’s primary purpose is facilitating decentralized computation. The latter has an immense scope, especially as it provides a user-centric and community-oriented infrastructure for Web3. 

However, Ethereum—or any other decentralized computing platform, for that matter — must be scalable to fulfil its role efficiently. It has to process thousands of transactions per second securely. Otherwise, DeFi applications, for instance, can’t compete with their Web2 counterparts like VISA.

On the contrary, Ethereum, like most other Layer-1 blockchains, currently has severe scalability limitations. The network clogs up due to traffic surges, as happened with Cryptokitties in 2017, DeFi in 2020, and NFTs in 2021. Congestion spikes gas fees, making Ethereum practically unusable for most users. This is a tremendous hurdle to mass adoption both for Ethereum and Web3. 

Ethereum’s recent transition to a Proof-of-Stake (PoS) consensus model — i.e., The Merge — is said to “set the stage for further scalability upgrades not possible under proof-of-work”. But these ‘upgrades’ cannot happen on Layer-1, for most parts, due to consequent security concerns. 

Also Read: The Merge is coming, but will it help Ethereum dominate the world?

Moreover, PoS doesn’t ensure scalability by itself. It requires implementing specific solutions like Sidechains, Plasma Chains, Channels, Validiums, and L2 Rollups. Among these alternatives, Rollups can leverage the underlying L1’s security effectively while increasing transaction speed and throughput at a much lower cost.

How do Optimistic and ZK Rollups work?

Generally, Rollups are smart contracts that relay transaction data and other information between L1s and L2s. The computation happens off-chain—i.e., on the L2—while the L1 only records and stores the hash or cryptographic address for completed transactions. 

L2 Rollups work by ‘rolling up’ or batching multiple transactions into a single block before sending them to the L1 for finalization. This reduces the burden on L1s, helping blockchains like Ethereum scale from roughly 15 TPS to 1000-4000 TPS. It also optimizes gas fees and lowers the overall costs for end-users, even during peak traffic. 

Optimistic and ZK Rollups adopt the core principle of relaying only critical, rolled-up smart contract data to the L1. However, they have key differences in their approach to validity and finality and the types of ‘proofs’ they use. 

As their name suggests, Optimistic Rollups assume all L2 transactions are valid unless someone challenges them. The process requires L1 nodes to submit ‘Fraud Proofs’ for potentially invalid transactions within a fixed timeframe. Unchallenged transactions are finalized when this window expires.

Most Optimistic Rollups, including Optimism and Arbitrum, are EVM-compatible, general-purpose smart contracts. According to Chris Dixon and Arianna Simpson of Andreessen Horowitz (a16z), “…this close adherence to Ethereum development paradigms results in a very easy transition for developers, wallets, and users: no new programming languages, minimal code changes to existing contracts required, and out-of-the-box support for the majority of existing Ethereum tooling.” 

ZK Rollups, on the other hand, adopt a more adversarial approach. Instead of assuming L2 transactions as valid, they use Zero-Knowledge Succinct Non-Interactive Argument of Knowledge or zkSNARK-based ‘Validity Proofs’ while publishing them on the L1. This method doesn’t require any challenge window and can achieve near-instant finality. It thus reduces the withdrawal time from days to hours, besides ensuring superior data privacy and integrity.

Also Read: The growing adoption of Ethereum in emerging markets

L2 challenges and the need for hybrid Rollups

Optimistic and ZK Rollups go a long way in scaling Ethereum, but they have limitations. For instance, there’s no guarantee that vigilant L1 nodes will successfully produce fraud proofs to refute invalid transactions published by Optimistic Rollups.

As for ZK Rollups, the development process is still highly complicated for most Web3 developers, which increases the scope for error and hampers contract-level security. Moreover, most existing ZK Rollups have difficulty achieving EVM compatibility, which restricts their accessibility. 

However, while Optimistic and ZK Rollups are still restricted in their individual capabilities, innovators like Yuan Su argue that the two approaches ‘can coexist and be unified’ to unlock their full potential. Further research has revealed the possibility of utilizing Optimism’s Cannon framework to integrate MIPS with ZK Proofs for effective EVM equivalence.

Innovating zkMIPS resolves the interoperability crisis facing ZK Rollups. It also eases the development process, enhancing accessibility for Web3 developers and project owners. But even besides that, the framework for Hybrid Rollups adapts a unique Optimistic Data Availability approach. 

During his talk at Devconnect AMS, Yuan explained Optimistic Data Availability as a method where it’s optimistically assumed that sequencers will make the data available for verification when required. This enables on-demand rollups which don’t have to roll up to the entire L1 data set all the time, drastically reducing availability costs and latency.

Given the capabilities discussed, Hybrid Rollups can arguably take Ethereum scalability to the next paradigm. Furthermore, open-source innovation and the consequent network effects can inspire other L1s to adopt Hybrid Rollups or similar solutions, ultimately making Web3 more scalable and cost-efficient.

Finally, as Web3 becomes faster, cheaper, and safer, it’ll be more relevant for diverse business cases and applications. This is the key to boosting adoption in the long run and in a sustainable manner. That’s the horizon for L2 Rollups, which can extend further with progressive 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

Join our e27 Telegram group, FB community, or like the e27 Facebook page

Image credit: Canva Pro

The post How layer-2 rollups boost Ethereum’s scalability for broader Web3 adoption appeared first on e27.