Hi, I'm Kyson (Zhu Jinhui) — a mobile engineer with 13 years in the field and a deep focus on iOS. Today I lead mobile at Sekai, a 0-to-1, million-DAU app built entirely in Flutter.

I graduated from Nanjing University of Information Science & Technology and have spent my career building mobile products at scale, including core apps at Alibaba (Ele.me) and Bilibili. Because I've often led features that span the whole stack, I've also picked up solid backend and Android experience along the way.

What I bring

Building for scale. The companies I've worked at — JD, Bilibili, Alibaba — are all public companies whose products serve tens of millions of users. At that scale, small problems become big ones fast, and solving them well demands real depth: understanding iOS internals, performance, and stability rather than just shipping features. That's the kind of rigor you mostly learn inside large, high-traffic systems.

Driving work across teams. Big organizations are made of many independent teams, each with its own priorities. Getting a good idea shipped means persuading other teams to invest in something that isn't on their roadmap — which only works if you frame the project so everyone benefits. Learning to lead through influence rather than authority has been one of the most valuable parts of my career.

Caring about process. Large projects live or die by their engineering and release discipline, and I've spent a lot of energy building the tooling and standards that keep big codebases healthy.

Selected work

Video player — Bilibili (International). I worked on the player team within Bilibili's international group — a 200+ person org with 60+ mobile engineers, targeting Southeast Asian users (a strategy similar to TikTok's regional editions). Video is a wonderfully hard domain: a clip travels through download → decoding → demuxing → rendering before it ever plays, and every stage hides challenges — caching, audio/video sync, instant start, and rebuffering. Our player kernel was built on ijkplayer (Bilibili's ffmpeg-based player). Much of the work was state management — for example, remembering exactly where a user left off when they switch apps mid-video — alongside features like caching, seeking, and variable-speed playback. The optimization work centered on reducing rebuffering: adapting to network conditions by pre-caching aggressively on good connections and stepping down resolution on poor ones.

Rider operations — Ele.me (0→1). At Alibaba I worked on Ele.me's local-services team, where I built a courier-management capability from the ground up. The hardest part of food delivery is the last leg — and our research showed new couriers were losing real time hunting for building numbers inside residential complexes, something map apps didn't cover. So we added in-complex building navigation to the Fengniao delivery app. It sounds simple, but the details were everything: sourcing accurate building data, letting couriers report errors so the data kept improving, building an algorithm to detect when a rider had actually reached a complex (and only then zooming in to show building-level navigation), and carefully gating the feature by device capability, since rendering building data is resource-heavy. I initiated this project and drove it to completion with both front-end and back-end teams — and it earned our team an award. Owning something end to end like that, from idea to launch, is the work I enjoy most.

On global products

I've always been drawn to internationalization. China's domestic internet has matured to the point where even small towns have food delivery — the room to grow at home is narrowing, which is part of why I moved toward global-facing work in the first place. I genuinely enjoy working across cultures, both the human side and the engineering side, and over the next few years I want to keep moving in a more global direction — ideally with a mature international team.

2023–2025: a self-directed chapter

After Bilibili, I spent about two years working independently. I built and ran a personal quantitative-trading system end to end — data pipeline, backtesting, automated execution — and took on freelance iOS and Flutter projects for clients. I also used the time to go deep on fundamentals I'd always wanted more of: algorithms, operating systems, and language internals. It was a period of exploration that ultimately clarified what I care about most — building high-performance mobile products — which is what drew me back to a full-time mobile leadership role at Sekai.

A note on Japanese

My university major was actually Japanese, and computer science was the hobby that quietly took over. My very first internship was building a map app for the Japanese market. After graduation I had to choose between the two paths; the internet was booming, so I chose code — but the language has stayed with me, and I'd love to put it to use again someday.