Hi,
I’m Hui Yao

> Creative AI Product Builder

My internship experience has centered on agent interaction strategies, model evaluation, and data analysis.

AT POIZON
I worked on the scenario strategy and conversational experience of an AI shopping assistant. By introducing scenario-based multi-turn guidance and more human-centered conversation strategies, I helped improve AI adoption and product exposure.
At Zhihuishu, I contributed to an AI knowledge-base product by building workflows for knowledge production and quality evaluation.

My interdisciplinary background in design and data analysis helps me understand how users interact with AI while validating product strategies through experiments and data.

EXPLORE MORE

Experience

# Education

Technical University of MunichInformatics, Exchange student
Shanghai Jiao Tong UniversityLCC, Master of Resources and Environment
Beijing Jiaotong UniversitySAAD, Bachelor of Environmental Design
MIT Online courseArtificial Intelligence: Scientific Data Analysis and Machine Learning Applications

# Internship Experience

NIO | AI Data Product Intern

  • Agent Operations Platform

    Own the development and iteration of Shuling’s internal operations platform. Integrate query data across Agent versions and business domains, and design AI tagging to assess whether answers meet expectations. Daily Feishu cards bring operations, product, engineering and evaluation teams together to resolve high-priority issues, supporting Agent iteration with shared data and workflows.

  • Agent System & Domain Skills

    Clarify dashboard metric definitions against business needs and iterate Shuling’s system and domain Skills. Design evaluation criteria and datasets to improve answer quality and drive product adoption.

POIZON | AI Product Intern

  • Intent Recognition & Conversation Strategy

    Built scenario-based, multi-turn guidance for an AI shopping assistant. Defined four stages of user intent to guide follow-up questions, explanations, recommendations and conversion, addressing unfocused recommendations and incomplete shopping journeys. Raised AI adoption by 4.47% and product-card exposure by 44.85% (relative gains).

  • Agent Tuning & Evaluation

    Independently designed four Agent personas and System Prompts, alongside an evaluation framework covering adoption, decision-making, interaction and recommendation. Increased overall AI adoption from 18.38% to 21.66%; decision-adoption UV and interaction-adoption UV improved by 3.38% and 1.50%, respectively (relative gains). Analyzed bad cases and user behavior to guide further iteration.

  • Reusable Skills for Data Analysis

    Packaged data retrieval and analysis into reusable Skills for an internal Agent. Automated metric monitoring, change-point analysis and weekly reporting, saving each team member about 2.5 hours per week and improving routine analysis and experiment reviews.

Zhihuishu.com | AI Product Intern

  • Multi-model Production & Evaluation

    Designed an automated medical knowledge-card pipeline and quality evaluation framework to address gaps in long-form generation and duplicate references. Coordinated models across parsing, generation, evidence verification and database ingestion, achieving a 97.8% ingestion acceptance rate. Analyzed 8,000+ bad cases and built an evaluation dataset based on the evidence-based medicine hierarchy to support production at scale.

  • User Research & Product Prototyping

    Researched domestic and international competitors to clarify positioning, audiences, use cases and business models for an AI knowledge base. Developed product directions for education and professional knowledge services, and completed three web-prototype iterations across conversational and detail pages to inform feature planning and commercialization.

# Project Experience

Improving Low-Resource Language Learning through Imbalanced Multilingual TrainingCore member
Food Sustainability Track TOP1 | AI-Driven Emotional Eating Companion SystemCore member
2025 SJTU Global Challenge Program | Southeast Asia waste plastic ProjectTeam Leader
National Innovation Training Project | Urban Wildlife Return PlanProject Leader
BJTU Art and Culture Center | Visual Arts DepartmentDeputy Head
BJTU SAAD Debate TeamCoach

# Entrepreneurial Experience

ZHU STUDIOFounder & Product Lead

Built an education services brand from the ground up for students pursuing recommendation-based graduate admission in China. Designed segmented offerings, tiered pricing, and delivery SOPs, transforming individual expertise into scalable courses and mentoring services. Achieved organic customer acquisition and paid conversion through WeChat, Xiaohongshu, and private communities, growing 1.6K+ followers, generating 50K+ likes and saves, and reaching RMB 300K+ in revenue.

# Honors Received

Gold Prize, Ecological Health and Sustainability Track, 12th Asian Design Award for Students

First Prize, 13th National Green Building Design Competition

National Second Prize, 7th Milan Design Week National College Outstanding Works Exhibition

Outstanding Work Award, 12th Annual Exhibition of Chinese College Artworks

Excellence Award, 12th National Green Building Design Competition

2023 National College Student Innovation Training Project Lead (Selection Rate: 4.54%)

Langji Global Talent Scholarship (2025)

Academic Excellence Scholarship (2021)

First Prize, Asia-Pacific Region, Global Sustainable Development Challenge

Track TOP1, FourC Challenge 24h Design Challenge

Team Leader, SJTU Global Challenge Program

Bronze Award, 4th Hong Kong Contemporary Design Awards

Outstanding Communist Youth League Member, University Level (2022, 2023)

Third Prize, University-level Innovation and Entrepreneurship Training Program (2022)

Third Prize, University-level Energy Conservation and Emission Reduction Competition (2023)

Urban Wildlife Return Plan 01. Urban Wildlife Return PlanNational Innovation Training Project AI-Driven Emotional Eating Companion System 02. AI-Driven Emotional Eating Companion SystemFood Sustainability Track TOP1 03. Emotional JournalMeituan AI Coding Project plasticwastedata.orgSoutheast Asia Plastic Waste Explorer 04. Southeast Asia Plastic Waste ExplorerIndependent Research Project
05To be continued… 05. To be continued…Future project

All projects2022—2026

Projects

01 / 04National Innovation Training Project

Urban Wildlife Return Plan

Project Leader
Back to projects

02 / 04Food Sustainability Track TOP1

AI-Driven Emotional Eating Companion System

Core memberIndependently built the product demo

PREPARING INTERACTIVE DEMO…

Interactive demo
Back to projects Open project · PDF

03 / 04Meituan AI Coding Project

Emotional Journal

Product design · Development · Deployment

PREPARING INTERACTIVE DEMO…

Interactive demo
Back to projects Open demo

04 / 04Independent Research Project

Southeast Asia Plastic Waste Explorer

Research · Data modelling · Visualization
Back to projects Open project

05 PHOTOGRAPHY

Fragments in Transit

A photographic notebook of places, weather, architecture, and moments in between.

ITALY

“Thence we came forth to rebehold the stars.”

0120

Preparing the photographs…

LET’S CONNECT

I’m always open to conversations about AI products, research
collaborations, and new opportunities.

Let’s build thoughtful AI products together！

Project

OPEN PDF ↗

Loading project…

HUI YAOPHOTOGRAPHY

ITALY01 / 20

USE ARROW KEYS TO BROWSE

© HUI YAO · PERSONAL VIEWING ONLY