<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Jieyi Deng — Writing</title><description>Data Scientist working across growth, product, machine learning and agentic analytics.</description><link>https://jieyi-deng.github.io/</link><language>en</language><item><title>More Than Prompting</title><link>https://jieyi-deng.github.io/writing/more-than-prompting/</link><guid isPermaLink="true">https://jieyi-deng.github.io/writing/more-than-prompting/</guid><description>SKILL.md turns a prompt into an operational protocol. What that looks like in advertising analytics, where a fluent answer built on the wrong metric is worse than no answer at all.</description><pubDate>Tue, 17 Feb 2026 00:00:00 GMT</pubDate><category>agents</category><category>skill design</category><category>ad analytics</category><category>llm</category></item><item><title>My 2025 with ChatGPT: What a &quot;Year in Review&quot; Doesn&apos;t Show</title><link>https://jieyi-deng.github.io/writing/my-2025-with-chatgpt/</link><guid isPermaLink="true">https://jieyi-deng.github.io/writing/my-2025-with-chatgpt/</guid><description>I exported a full year of ChatGPT history and analysed it. The conversation tree holds more than the official recap shows — including three different URL fields that reveal how retrieved sources actually reach an answer.</description><pubDate>Tue, 30 Dec 2025 00:00:00 GMT</pubDate><category>data analysis</category><category>llm</category><category>personal analytics</category></item></channel></rss>