Conversation with Sea co-founder and Shopee Chief Product Officer David Chen.
Our Executive Function series features the perspectives of industry leaders driving the AI transformation.
Sea Limited (Sea) is a Singapore-based global technology company with businesses spanning digital entertainment, e-commerce, and digital financial services. Its engineering teams build and operate products at scale across some of the world’s most dynamic markets.
Sea is rolling out Codex across its developer organization, and our internal data shows that 87% of users are weekly active users. For the company, AI-assisted software development is not just about marginal productivity gains; it represents a deeper shift in how engineering teams manage complexity, build resilient systems, and move from idea to implementation.
We spoke with Sea co-founder and Shopee Chief Product Officer David Chen about why the company is making this bet, how AI agents are changing the way developers work, and what AI-native software development means for Southeast Asia and the broader Asia-Pacific region.
David, Sea operates across some of Southeast Asia’s most dynamic markets. Why did you decide to roll out Codex across the entire engineering organization? What stood out to you about the product?
At a company like Sea, engineering is not just about writing code; it is about managing system complexity at scale across fragmented, highly localized markets. We believe the continued evolution of AI is driving a fundamental shift in how software is created and how engineering teams operate at scale.
Agentic AI coding tools like Codex are more than a local productivity boost. They act as a structural multiplier, helping our engineering organization increase speed, responsiveness, and efficiency in an increasingly complex operating environment.
What specifically stands out about Codex?
What stands out most is that it goes beyond autocomplete and delivers deep context awareness across our large, distributed codebase. In a microservices architecture at scale, the real friction is not typing syntax; it is tracing dependencies, understanding legacy logic, and maintaining reliability under peak load.
Codex acts like a localized knowledge engine, dramatically reducing the time engineers spend navigating unfamiliar services and allowing teams to shift cognitive load to higher-value work such as architecture and product innovation.
Internal feedback suggests that people are using Codex actively for code understanding, debugging, and feature development. What does that reveal about how developers are using Codex day to day? And how are AI agents beginning to change the way Sea builds software?
It has been encouraging to see the adoption trends for Codex across our developers, especially the growth in frequent users. Many have said it improves the speed of experimentation and their development workflow. Based on internal feedback, 73% of developers who rated Codex 4 or 5 stars said they would recommend it to a colleague.
The most profound change is that our developers are using Codex to think better, not just type faster. We are actively moving from treating AI as a passive autocomplete mechanism to integrating it into agentic workflows.
In practice, this means AI agents are increasingly operating within our CI/CD pipelines—reasoning about product requirements, autonomously proposing test-driven implementation approaches, identifying edge cases in distributed systems, and accelerating debugging cycles.
A lot of people assume AI is only about speed. At Sea, we are also using it to strengthen engineering discipline. By having AI rapidly prototype alternative implementations and generate comprehensive test coverage, we are moving faster while systematically paying down technical debt and delivering more resilient systems.