Rebuilding My LinkedIn for a Global Audience — as an Engineer, with a Browser Agent
Why I dropped the Chinese version of my profile, how I rewrote the headline and About section, and how a browser agent handled the 30+ edit steps for me.
Most of my career has happened in Taiwan, and my LinkedIn profile looked like it: a Chinese primary profile with an English secondary version bolted on, both half-maintained. When I decided to build TripPro for Asian English-speaking travelers, that profile stopped making sense. The people I want to reach — collaborators, early users, other engineers working on social products — read English first.
This post is the “hello” of this blog. It covers the decisions behind the rewrite, and the small automation project that came out of it: using a browser agent to apply the changes instead of clicking through LinkedIn’s editor by hand.
Why I dropped the Chinese version
Two versions of a profile sound like reach. In practice they were drift. Every time I updated one, the other went stale, and LinkedIn picks which one to show based on the viewer’s language setting, so I never really knew what people were seeing.
Keeping only English is a positioning decision, not just a maintenance one. It says who the profile is for.
Rewriting the headline
The old headline was a job title. The new one is a sentence about what I build and for whom: social sharing × social tech, based in Taiwan, currently building TripPro. Job titles tell people where you are; a sentence tells them why they should care.
Rewriting the About section
I structured it as: what I work on now, what I’ve built before, and how to reach me. Short paragraphs, no buzzword lists. The goal is that someone landing from a blog post or a dev.to article gets the same story here.
Letting a browser agent do the clicking
LinkedIn’s profile editor is a lot of small modals — headline, About, each experience entry, skills, secondary languages. Editing everything by hand is 30+ steps of open, clear, paste, save. I ran the edits through ego-browser, an AI-agent-first Chromium browser I’ve been working on, which lets an agent work in its own task space while reusing my logged-in session.
I’ll go deeper on the browser side in a later post. The short version: the agent worked through the edit list, took a screenshot after each save, and I reviewed the results at the end instead of babysitting each modal.
What’s next
The next posts will get more technical: importing social links without an API, and why we built an agent-first browser in the first place. If any of that sounds relevant to what you’re working on, say hi on LinkedIn or by email.