yipingz.

Problem solver • Engineer • ex-Factory Worker • ex-Gamer

I love solving interesting problems and exploring the limits of AI. These days, I'm focused on recommendation systems — what they can become in the AI era. 喜欢解决有意思的问题，探索 AI 的边界。目前专注推荐系统，探索它在 AI 时代的可能性。

Join the newsletter

Scroll to inspect ↓

02 / FIELDS

Current areas
of work.

THINK.01 / ASSUMPTIONS

Problem
framing

Making assumptions explicit, then checking them against data, constraints, and observed behavior.把假设说清楚，再用数据、约束和实际行为检验。

INTEL.02 / RECOMMENDATION

Recommendation
systems

Retrieval, ranking, generative approaches, and the evaluation choices that connect them to product outcomes.关注召回、排序、生成式方法，以及与产品结果相连的评估方式。

MAKE.03 / PRACTICE

Building &
operating

What changes between a prototype and production: latency, feedback loops, incentives, and maintenance.记录从原型到生产环境之间的延迟、反馈回路、激励和维护问题。

03 / WRITING

Draft topics.

DRAFT / 001

Starting with the system’s assumptions

How goals, data, and constraints shape model behavior before training starts.

这些题目还在整理中，完成后会在这里发布。

yipingz.

Practitioner's notes on AI & recsys.

Current areasof work.

Problemframing