EVIE / YUXIN HOU · 侯煜欣
I’m Evie Hou, a Sociology PhD student and former startup CTO. I study social opinions and survey simulation, and develop AI learning tools.
Sociology PhD, Peking University · Visiting Student Research Collaborator, Princeton
Research and projectsAn AI portfolio assistant, grounded in public sources.
Curated answers, not live AI. Your question stays in the browser.
02 / About
My experience includes startup R&D, education nonprofits, a UNICEF internship, and product and AI consulting, alongside my academic training.
I earned my B.E. in Mechanical Engineering and M.A. in Education at Tsinghua, then joined Peking University’s Center for Social Research in September 2024. I’m now a third-year Sociology PhD student, with a one-year visiting appointment at Princeton.
I taught myself programming, developed platforms and algorithms, and led an R&D team of more than 20 people as a startup CTO from 2016 to 2022.
AboutFollow the stops · hover, focus or tap to explore.
Later earned a B.E. in Mechanical Engineering and an M.A. in Education.
Started a company and served as CTO. Learned programming independently, developed multiple platforms, designed algorithms, and led an R&D team of more than 20 people. Left the company in 2022.
A separate NGO internship within the shared 2020–2023 period. Across the two NGO internships, work included data collection, technical support, program design, and product design; individual dates were not supplied.
A separate NGO internship within the shared 2020–2023 period. Across the two NGO internships, work included data collection, technical support, program design, and product design; individual dates were not supplied.
Led an SEL video resource program during the internship.
Sociology PhD student at the Center for Social Research; currently in the third year.
Worked as a product and AI consultant for three companies.
One-year Visiting Student Research Collaborator (VSRC).
Two qualitative bands supplied by Evie: strengths and areas with less experience.
03 / Research
Survey simulation, opinion correlation, and opinion dynamics.
Survey simulation
An LLM agent fits and evaluates generators of synthetic survey data.
Research code and experiment reports; includes empirical-copula transfer and checks for record copying.
Comparative opinion research
Compares associations among nine policy opinions in surveys and social media across the US, Europe, and China.
Pairwise correlations, semantic similarity, and effective dimensionality. Public analysis code.
Research
Opinion dynamics on social media.
Research area; project methods and results are not listed here.
04 / Projects
Learning applications and agent software. Each entry identifies its available implementation and release status.
Web application
An AI learning platform for reading, writing, courses, and research projects, with structured thinking cards and process records.
Public website and application entry point are live. English product branding: The Mark of Thinking.
Desktop application
A course-learning application with streamed tutoring dialogue, local Python agent execution, and synchronized course progress.
Public desktop releases; Electron frontend, Python sidecar, and FastAPI backend.
Prototype
An agent runtime with model-provider adapters, tool execution, session memory, and a chat API.
Public prototype code. Organizational messaging integrations are architecture plans, not verified deployed features.
Project
Project description and release status are not public on this page.
Project
Project description and release status are not public on this page.
A repository reading note on statistical fidelity, record copying, and feasible versus oracle conditions in SSDataAgent.