
Main takeaways:
A new generation of markets is forming around capital, compute, intelligence, and data. Capital moves onto programmable rails, compute becomes strategic input, intelligence becomes abundant and autonomous, and high-quality data becomes increasingly valuable. AI agents are beginning to take action, transact, and coordinate. And as AI moves deeper into science, evidence and accountability are becoming as important as model capability.
As these forces converge, new markets are needed to allocate resources, coordinate autonomous systems, verify outcomes, and establish trust.
That is where EASY Residency Season 5 is focused.
“The big opportunities emerge when a new capability creates a new constraint: programmable capital needs deeper market structure; fragmented compute need reliable allocation; autonomous agents need permission and accountability; and physical and regulated AI need proprietary data and workflow integration. At YZi Labs, we want to back founders who identify these bottlenecks early and have the depth and conviction to turn them into enduring companies.”
— Ella Zhang, Managing Partner, YZi Labs
What We’re Looking For in Season 5
[Continue from Season 4] Stablecoin Payments, FX & Treasury Infrastructure
Stablecoins are turning internet-native money into practical infrastructure for cross-border settlement, merchant payments, payroll, remittances, treasury, and machine-to-machine commerce.
We are continuously interested in cross-border B2B settlement, stablecoin-native FX, treasury management, merchant acquiring, payroll, remittances, liquidity orchestration, reconciliation, compliance, and programmable controls for enterprises and fintechs. We are also interested in agent wallets, delegated spending, usage-based payments, and settlement rails for autonomous software.
Exchange-Grade Onchain Markets
As onchain trading approaches centralized exchanges in performance and usability, we see two complementary opportunities.
The strongest teams will either own a critical execution or risk layer, or use privileged distribution and liquidity to create financial products that could not exist within a closed exchange environment.
DeFi Application Layer for Tokenized Equities & RWAs
As tokenized blue-chip equities reach meaningful trading volume, much of the market infrastructure remains focused on issuance, distribution, and secondary trading. The broader financial application layer—including credit, collateral management, risk transfer, and structured products—remains underdeveloped.
We are interested in protocols that enable tokenized assets to become productive onchain collateral and composable financial building blocks. Tokenized assets create new collateral and settlement possibilities, but teams must solve the hard edges: market hours, corporate actions, legal claims, custody and redemption, liquidity, pricing, and risk management.Promising wedges include RWA-collateralized lending, structured yield products built on dividend-bearing assets, and capital-efficiency infrastructure that deepens liquidity across tokenized asset markets.
Institutional-Grade Onchain Options
Options and other risk-transfer products remain underbuilt for tokenized assets, particularly where users need reliable margin, settlement, and liquidity across fragmented venues.
We see early demand from asset holders seeking hedging and yield strategies, but sustainable product-market fit will depend on market-maker participation, collateral efficiency, reliable settlement, and repeat institutional usage.
We are interested in teams building the market structure required for repeat institutional usage, including robust margining, reliable settlement, market-maker tooling, and capital-efficient liquidity mechanisms.
Privacy Infrastructure
Privacy is a prerequisite for bringing sensitive financial activity onchain. Individuals and institutions cannot expose balances, counterparties, positions, or commercial terms on public ledgers, yet they must still satisfy compliance, audit, and risk-management requirements.
We are interested in privacy infrastructure that enables selective disclosure rather than absolute anonymity. Promising wedges include confidential DeFi, private stablecoin settlement, programmable compliance, privacy-preserving identity and credentials, and confidential payroll, vendor, treasury, and institutional trading workflows. The strongest solutions will combine robust cryptography with practical key management, regulatory interoperability, and a user experience that can support real financial activity at scale.
Post-Quantum Migration & Cryptographic Agility
As the industry prepares for a post-quantum future, we are interested in infrastructure that enables wallets, custodians, exchanges, and protocols to upgrade safely and incrementally. Promising wedges include crypto-agile wallet and custody SDKs, hybrid signing, hardware-wallet upgrades, post-quantum threshold and MPC systems, account abstraction and key rotation, migration orchestration, and efficient onchain verification of post-quantum signatures. The strongest solutions will integrate with existing products and distribution, support multiple cryptographic standards, and provide a practical migration path for accounts and assets already onchain.
Compute Market Infrastructure
As AI compute fragments across hyperscalers, neoclouds, regional operators, and thousands of emerging microclouds, buyers need more than another GPU listing page.
We are interested in companies building the procurement, routing, reliability, and benchmarking layers that make heterogeneous compute usable at scale.
In particular, we see opportunities in inference-specific marketplaces that optimize for latency, SLA, throughput, and cost across microclouds, as well as regional / non-US compute access layers in the Middle East, SEA, India, Europe, Japan, and Korea, where supply remains fragmented and buyers still rely heavily on local neoclouds. The strongest opportunities will own workflow or risk, not just supply aggregation.
Compute Capacity Markets
As GPU capacity becomes a volatile strategic input, AI companies and infrastructure providers need better ways to secure supply, manage price risk, and finance capacity.
We are interested in standardized contracts, verifiable delivery and SLAs, capacity financing, forward procurement, and hedging markets grounded in real compute demand—not speculative exposure.
Trusted Agent Infrastructure
OpenAI slowing model development over cyber-capability risks highlights a new bottleneck for AI adoption: not whether agents can act, but whether their actions can be trusted, constrained, and audited.
As agents gain access to data, tools, and capital, the bottleneck shifts from capability alone to identity, authorization, containment, auditability, and recovery.
We are interested in agent identity, delegated authority, least-privilege permissions, secure tool access, policy enforcement, tamper-evident logs, monitoring, rollback, and agent-security evaluation. Onchain or cryptographic verification is especially valuable where it creates a measurable trust, settlement, or coordination advantage.
Physical AI Data Infrastructure
Robotics bottleneck is no longer just model architecture, but real-world data, simulation, annotation, deployment telemetry, and eval loops. Strong opportunities include egocentric data collection, synthetic-to-real pipelines, robot evals, fleet learning tools, and incentive networks that coordinate distributed contributors to capture diverse real-world embodied data.
The moat comes from verified data provenance, mechanisms to reward useful data and filter low-quality submissions, and deep connections between robotics/model companies and scalable data supply.
As AI assistants become proactive, multimodal, and persistent, interaction will extend beyond the phone into voice, wearables, ambient devices, and intent-based interfaces. We are interested in the memory, privacy, and action layers that make these interfaces useful and trusted.
Privacy-First Memory & Action Interfaces
As AI assistants move from chatbots to always-on companions across phones, speakers, earbuds, glasses, and workplace tools, the core bottleneck becomes persistent personal context: what the user has seen, heard, decided, bought, prefers, and wants to remember.
We are interested in privacy-first memory layers that can capture, structure, permission, and retrieve user context across devices and agents, starting from narrow wedges such as meeting memory, personal CRM, shopping memory, health/lifestyle memory, or family/home memory. The long-term opportunity is becoming the user-owned context layer for all AI interfaces.
Voice-First AI Interface
Voice is emerging as an important mass-market AI interface beyond the phone screen because it is natural, low-friction, and hardware-light compared with AI glasses or BCI.
We are interested in voice-first applications and infrastructure that turn spoken intent into actions, memory, transactions, and workflows, especially in high-frequency contexts such as meetings, customer calls, family coordination, field work, learning, and commerce. The winning products will not be generic voice chatbots, but vertical voice agents with workflow ownership and measurable ROI.
AI in Regulated Healthcare Operations
As hospital operating margins shrink while net revenue leakage grows, health system budgets are rapidly migrating into AI solutions that offer immediate ROI. We are interested in payer-side prior authorization and claims review, revenue integrity and coding tools, specialty-specific workflow agents, and liability layers that resolve accountability.
The Lab Data Middleware Layer
The primary constraint on AI in wet biology is no longer model architecture or robot hardware, but the absence of a clean, machine-readable experimental data layer between instruments and models. While capital floods adjacent materials-science mega-rounds, the wet-lab middleware layer remains conspicuously underfunded. We are interested in instrument APIs, protocol execution provenance, automated run QC, negative-result capture, and closed-loop scheduling that turn robotic labs into data-generating assets. The strongest opportunities will embed directly into the physical execution path to build a persistent, proprietary data feedback loop.
AI-Native Clinical Development and Regulatory Submission
Pharma and CROs present a clean buyer structure in healthcare, offering large deal sizes and short sales cycles without EHR or reimbursement risks—with value concentrated in execution speed given zero AI-designed drug approvals to date. We are interested in protocol optimization, site and patient matching, regulatory submission QC against emerging FDA/EMA guidelines, trial data standardization, and real-world evidence generation.
Bioprocess, Scale-Up and CDMO Intelligence
Manufacturing scale-up remains the most economically neglected bottleneck in bio relative to its market weight, offering clear ROI driven by the high cost of failed batches. We are interested in bioreactor process optimization, tech-transfer prediction, batch failure prediction, yield and titer optimization, and CDMO capacity planning.
Web3 × Bio — Consent, Provenance and Payment Rails for Human Biological Data
Protecting and permissioning human genomic data requires infrastructure that survives changes of corporate control such as corporate bankruptcy. We are interested in cryptographic consent receipts, revocable and auditable authorization, dataset provenance attestation, and contribution-linked payment rails. These opportunities can sell to biobanks, health systems, pharma data buyers, and regulators as enterprise compliance infrastructure revenue, rather than pure consumer-facing tokenomics.
How the Residency Works
The program runs for 10 weeks, structured as 5 weeks online plus 5 weeks in person in Thailand.
What founders get covered
1. Investment
EASY Residency founders may receive up to $500,000, structured as:
In addition, all founders admitted to EASY Residency will also be eligible for up to $2 million equivalent token investment via an uncapped SAFE from a leading token aggregation platform building integrations across major foundation models.
2. Infrastructure for builders
3. Hands-on acceleration
How to apply
Applications are now open and will close at 23:59 (GMT-7) on September 13.
Apply here: bit.ly/45WOv3M
We’ve upgraded our founder portal to make the EASY Residency application process smoother and easier.
If you’ve applied through our previous questionnaire, there’s no need to apply again — your application is already in our system, and we’ll reach out directly if we’d like to move forward.
Hope to meet you in Thailand.
About YZi Labs
YZi Labs manages over $10 billion in assets globally. Our investment philosophy emphasizes impact first—we believe that meaningful returns will naturally follow. We invest in ventures at every stage, prioritizing those with solid fundamentals in Web3, AI, and biotech.
YZi Labs' portfolio covers over 300 projects from over 25 countries across six continents. Notable portfolio companies include Trustwallet, CoinMarketCap, Polygon, Injective, Ethena, Safepal, Better Payment Network, Aster, XAI, and more. More than 65 of YZi Labs' portfolio companies have gone through our incubation programs. For more information, follow YZi Labs on X.
About EASY Residency
EASY Residency is YZi Labs' flagship global incubation program supporting early-stage, long-term founders across Web3, AI, and biotech. The program fosters a community where builders find peers, mentors, and investors who share their conviction.
Each season brings together a curated group of startups from around the world to live and build side by side, culminating in a high-profile Demo Day with leading investors and industry leaders.
Disclaimer
The information provided in this article is intended for informational purposes only and does not constitute investment advice, endorsement, analysis, or recommendations with respect to any financial instruments, investments, or issuers. [Regulatory/framework references to be added once the Thailand-specific claims in this doc are finalized and sourced.] Founders should consult their own legal, tax, and financial advisors for current applicability. This article may contain forward-looking statements which are by nature subject to risks and uncertainties. Investment in cryptocurrency and DeFi projects involves substantial risk, including the risk of complete loss.