EN Eason Ni

AI Search / Recommendation / Product Intelligence

Eason Ni

I turn messy intent, product signals, and research questions into search and recommendation systems people can actually use.

Query understanding Retrieval quality Ranking judgment Reusable workflows
View Focus Find Me
About

A practical research home base for AI-powered discovery.

My work sits between user understanding, item understanding, matching, ranking, and product judgment. I care about systems that improve real workflows, not isolated demos that look clever once.

01

Business questions first

Start from the decision, then map data, model, evaluation, and UX.

02

Evidence over slogans

Keep source notes, bad cases, and reusable checklists close to the work.

Selected Work

Three directions I keep building around.

Each direction combines research, product analysis, and small runnable workflows.

01

AI Search

Query understanding, multimodal retrieval, ranking evaluation, and search-result quality loops.

02

Recommendation

User intent, content and product signals, distribution mechanics, and recommendation product judgment.

03

Applied Research

Turning research questions into structured notes, project plans, demos, and decision-ready summaries.

Method

The operating loop

A lightweight loop for moving from ambiguous product questions to reusable knowledge.

  1. Explore Clarify the intent

    Define the user need, product context, and evidence needed to judge quality.

  2. Build Connect signals

    Map queries, items, content, models, and evaluation into a system view.

  3. Distill Make it reusable

    Keep the output readable: notes, demos, checklists, and next-step decisions.

Notes

Notes that make the work easier to reopen.

This site is a public index for research notes, project summaries, and small demos around AI search, recommendation, product intelligence, and agent workflows.

AI search and retrieval quality Recommendation system product judgment Applied research workflows
Contact

The fastest way to find me.

Reach out if the topic is AI search, recommendation systems, product understanding, applied research, or practical knowledge workflows.

GitHub: Eason-Ni WeChat Official Account: 学习的一天天
WeChat Official Account QR code for 学习的一天天
Scan to follow my WeChat official account.