Xudong Han Verified by creator
@Xudong07452910 X
A PhD student writing on AI agents: 10K in 108 days
A UK PhD student in LLMs and AI agents began posting in Chinese on X in May. He says his first article went out when he had 4 followers; daily paper breakdowns followed, and recent median views are 11× his start.
- Report
- v1
- Data as of
- 2026-11-14
- Sample
- 200 days of posting counts
Xudong Han@Xudong07452910Article: Building a personal agent in 2026: how I used vibe coding and borrowed open-source parts to build a work system I use every day, in 15 days (the method)
View on X →At a glance
- What
- An X account where a PhD student in the UK breaks down AI agent papers and open-source projects in Chinese.
- For whom
- Chinese-speaking developers, product managers and students working on AI agents.
- Money
- None visible.
- How it grew
- On day 8 of posting, a method article on building your own agent got 160K views; then daily paper breakdowns, and he says 10K in 108 days.
X followers Fact
15.7K
posting since 2026-04-29
Build-your-own-agent method Fact
160K
views, 1,312 bookmarks
To 10K followers Self-reported
108 days
May 1 → Aug 17
Median views Fact
460 → 5,123
start → recent
Summary
The play
Two opening articles on how he built his own personal agent established him as “the PhD who knows agents”; then daily Chinese breakdowns of new agent papers and open-source projects, plus lots of replies and reposts of good work from small accounts.
Can you copy it? Our read
What you can copy is the structure: two opening articles on a system you really use, then a daily breakdown of new research in one niche. But he has the standing of a PhD student in the field and had reposts at the start; the follower path is self-reported, and no monetization is visible.
- What it takes
- Real understanding of a niche, and the ability to read papers and write clearly.
- Startup cost
- From May to July, about 9 posts and 34 replies a day.
- Time to a verdict
- He says 11 days to 1,000 followers and 108 to 10K.
- Window
- AI agents are still hot; the barrier for paper-breakdown accounts is expertise, not timing.
Key findings
Two articles on how he built his own agent
The May 6 method article got 160K views, about 347× the median views of his first two weeks. See play 2.1 (subscribers)
Daily breakdowns of papers and open source
808 posts and 3,122 replies from May to July; recent median views are 11× his start. See play 2.2 (subscribers)
Reposting good work from small accounts
He says reposts lifted him, and from 6,000 followers he reposts a few small accounts’ good posts every day. See play 2.3 (subscribers)
What this channel is
- Main platform
- X
- Format
- Short breakdowns of papers and open-source projects, plus a few articles
- Who
- PhD student at the University of Sussex, working on LLMs and AI agents (X bio)
- Started
- Joined 2020-11, started posting 2026-04-29
- Team
- No team seen
- Region
- London, UK (X profile)
Every number is tagged
- Fact straight from public data
- Derived calculated from facts
- Estimate from published pay ranges
- Self-reported what the creator says
- Our read our judgement
01How it happened
How the X following grew
- X 15.7K
- Fact
- Self-reported
X doesn’t publish follower history. The points in between are all from his own posts (self-reported); he posted at nearly every round number. November 14 is measured.
- under 1×
- 1–3×
- 3–10×
- 10× and up
Each bar is one day, in date order; its height is how many main posts (originals, quotes and articles; not replies) went out that day. This chart counts posts per day rather than listing each post. May 2–3, 2026 read zero for every account we track: a gap in the data.
Is this your account? Claim it