# tanghoong

> Levelling up in public — notes from building AI systems solo.
> Written by Charlie Tang Hoong. Handle: @charlie.
> 3 articles, some published in both English and 中文.
> Every post is a markdown file in a public repository:
> https://github.com/tanghoong/github-markdown-blog/tree/main/contents

## Posts

- [Sync or async for an AI support chatbot](https://blog.tanghoong.com/posts/engineering/sync-or-async-for-an-ai-support-chatbot/) (2026-08-02, engineering): If customer A is waiting for the LLM, does customer B have to wait too? The answer is an architecture decision, and it matters long before you have a thousand customers.
- [An AI agent does not need to remember everything](https://blog.tanghoong.com/posts/engineering/an-ai-agent-does-not-need-to-remember-everything/) (2026-08-01, engineering): Producing a convincing answer is no longer the hard part. The hard part is the agent still working tomorrow — and that turns out to be a state problem, not a memory problem.
- [Why this blog is a folder of markdown files](https://blog.tanghoong.com/posts/welcome/) (2026-07-29): No database, no CMS, no admin panel, no server. Not because those things are bad, but because every one of them is a cost I would keep paying long after the novelty of setting it up wore off.

## 中文文章

- [AI 客服系統該用同步還是非同步？](https://blog.tanghoong.com/posts/engineering/sync-or-async-for-an-ai-support-chatbot-zh/) (2026-08-02, engineering): 客戶 A 在等 LLM 回應的時候，客戶 B 要不要跟著等？這是個架構問題，而且遠在你有一千個客戶之前就開始有影響。
- [AI Agent 不需要記住所有事](https://blog.tanghoong.com/posts/engineering/an-ai-agent-does-not-need-to-remember-everything-zh/) (2026-08-02, engineering): 讓 agent 講出一個像樣的答案，已經不是最難的部分。難的是它明天還能用——而那其實不是記憶問題，是狀態問題。

## Translations

- AI 客服系統該用同步還是非同步？ is the 中文 edition of "Sync or async for an AI support chatbot" — the same article, not a second one.
- AI Agent 不需要記住所有事 is the 中文 edition of "An AI agent does not need to remember everything" — the same article, not a second one.

## Reference

- [Archive](https://blog.tanghoong.com/archive/): every post grouped by year
- [RSS](https://blog.tanghoong.com/rss.xml): full feed
- [JSON Feed](https://blog.tanghoong.com/feed.json): the same, as JSON
- [Sitemap](https://blog.tanghoong.com/sitemap-index.xml): every page