The blockchain ecosystem has evolved into a fragmented landscape of competing networks, each with its own protocols, gas tokens, and development quirks. For developers and users alike, navigating this multi-chain reality has become increasingly complex. But what if building across multiple blockchains could be as simple as deploying a traditional web application?
Shun Kakinoki, Head of Cross-Chain at Sequence, has been tackling this challenge head-on. After founding LightDotSo, a chain abstraction protocol that raised $500K in venture funding before being acquired by Sequence in 2025, Shun has positioned himself at the forefront of efforts to unify the fragmented blockchain landscape. His vision: making multi-chain development as straightforward as single-chain development, where developers “write once and deploy everywhere.”
Coming from a background in deep learning and mobile/web development, Shun brings a unique perspective to blockchain infrastructure. His experience in Silicon Valley during his formative years, combined with his technical expertise across traditional and blockchain development, has shaped his approach to solving one of web3’s most persistent problems.
Shun breaks down for us the technical challenges of multi-chain development, explains how chain abstraction works in practice, and shares why the recent combination of chain abstraction with account abstraction could finally deliver the seamless user experience that web3 has been promising. Whether you’re a developer considering multi-chain architecture or a business exploring blockchain applications, his insights offer a roadmap for navigating the increasingly complex world of decentralized infrastructure.
What inspired you to focus specifically on chain abstraction, and why is this problem so critical for the future of web3?
I focused on chain abstraction because fragmentation was the biggest barrier I saw – users and developers shouldn’t have to care what chain they’re on. As more and more chains come online; the fragmentation problem worsens. Solving this unlocks a true “internet of value” where apps can route transactions seamlessly.
The term “chain abstraction” is still relatively new to many developers. How do you explain it in practical terms, and what does it mean for the average developer building blockchain applications today?
Practically, chain abstraction means developers build once and their logic runs everywhere; the system chooses the right chain and business logic behind the scenes. For developers, it’s like moving from just focusing on the core code to then forgetting about the whole process towards deployment..
Traditional web development has standardized protocols and infrastructure. What are the biggest technical limitations developers face when building across multiple blockchains, and how does your approach solve these challenges?
The biggest limitation is each blockchain having different RPC quirks + preparation, gas tokens, and bridging mechanisms. Our approach standardizes those into one interface, so multi-chain is as easy as single-chain.
Can you walk us through what unifying all EVM chains actually looks like from a developer’s perspective? What changes in their workflow?
Unifying EVM chains means developers no longer write chain-specific boilerplate. They just call functions, and the infrastructure decides whether it settles to any L2 like Arbitrum, Base, or Optimism, or any L1 like Monad, Berachain or Polygon.
For businesses and developers just starting to explore multi-chain development, what are the most common mistakes you see, and what practical steps should they take to avoid fragmentation issues?
The most common mistake is hard-coding to a single chain, which creates lock-in. Start with abstraction layers early so your app can scale cross-chain without rewrites so that “write once, works anywhere” mantra of web2 works for the same extent and ease as web2.0 code.
The acquisition by Sequence combines your chain abstraction expertise with their account abstraction and infrastructure capabilities. How does this change what’s possible for web3 applications, especially in terms of user experience?
With Sequence, chain abstraction plus account abstraction means users can log in once and pay anywhere – apps can deploy and write code to any given chain but accept payments from all chains just like how web apps work across any browsers today. It turns crypto UX from “multi-wallet chaos” into one seamless experience.
Coming from a background in deep learning and mobile/web development before entering blockchain, what lessons from traditional software architecture apply (or don’t apply) when designing for a multi-chain future?
From deep learning and the web, I learned the value of modular layers – clean abstractions and shared APIs. In multi-chain, the same applies: abstraction is the only way to manage complexity at scale.
Security and interoperability often seem at odds with each other in blockchain development. How do you balance the need for seamless cross-chain functionality while maintaining the security guarantees that make blockchain valuable?
Security and interoperability balance comes from strict verification at boundaries: cryptographic proofs, audited bridges, and minimized trust assumptions. We never compromise security for convenience, but we design workflows that feel seamless.
What emerging trends in blockchain infrastructure excite you most, and what advice would you give to developers who want to build applications that can scale across the entire web3 ecosystem?
I’m most excited about intent-based infrastructure, where users just express what they want and the system figures out the best execution path. My advice: design for scale now, assume your app will live across many chains.
For technical teams considering whether to build for multiple chains from the start versus focusing on one blockchain, what factors should guide that decision, and how has the tooling evolved to make multi-chain development more accessible?
Teams should consider whether their app is user-facing or infra-heavy – consumer apps almost always need multi-chain reach, infra may start narrow. Tooling has matured so much that building a multi-chain app is no longer 10x harder; it’s often just a config choice and one single SDK installation away from accepting payments everywhere.
