YH

Where art
meets intelligence.

I turn messy enterprise workflows into AI systems people actually use. The aim is never to replace their judgment, but to give it better tools: scoped together, explainable by design.

Portrait of Yuqing in a classical style

Human-centered intelligence

I make AI systems more human, explainable, and useful.

I started in mechanical engineering, where a design only counts if it works in the real world. I've kept that standard in AI. Today I build enterprise AI at Arcas Agentic, taking pilots in legal, finance, and deal advisory from first client conversation to something a team runs on their own. Outside work I've taught dance for eight years, which is where I learned to explain hard things simply.

Product-minded

Turning requirements into usable, operable systems.

Explainable

Designing AI that people can inspect and trust.

Human-centered

Guided by Data Humanism and real-world context.

End-to-end

From the first scoping call to the shipped release.

  • LangChain
  • LangGraph
  • RAG
  • MCP
  • Multi-agent systems
  • Document intelligence
  • LLM evaluation
  • PII masking
  • Microsoft Foundry
  • Azure
  • Python
  • TypeScript
  • React
  • Next.js
  • Vue.js
  • FastAPI
  • PostgreSQL
  • Supabase
  • Docker

Technology needs judgment,
not just execution.

I connect research, engineering, and product thinking to build intelligent systems that people can understand and use.

Arcas Agentic · Dec 2025 to now

Enterprise AI pilots and agentic workflows

End-to-end AI pilots across legal, finance, and deal advisory: document intelligence pipelines, RAG-based agents, and LLM extraction for contracts, IMs, and valuation materials.

Full-stack developer, solo · Sep to Dec 2025

Dance studio member management system

A web app replacing manual studio admin: QR code check-in with real-time capacity tracking, automated subscription deductions, and a self-service member portal. 19 releases shipped on Vercel with Next.js and Supabase.

Frontend developer, team of five · Apr 2024 to Jan 2025

Group music recommendation platform

A collaborative music player with Spotify integration, group song voting, and explainable real-time recommendations. Processed 1M+ song audio features, trained a LightFM model, and built the Vue interface from Figma designs.

Master thesis · University of Zurich · 2025

Explainable AI for fashion recommender systems

A fashion recommender that explains itself, integrating graph neural networks, interactive visualization, and Data Humanism philosophy.

Dec 2025 – now · Zurich

AI Solutions Engineer, Arcas Agentic

Building the AI-native Arcas platform and delivering enterprise AI pilots, from scoping client requirements to production-ready workflows.

Oct 2025 – now

Software Engineer, bloem

Full-stack development of a circular fashion marketplace in TypeScript and PostgreSQL.

Jul 2021 – Jun 2022 · Shanghai

Project Development Engineering Intern, Apex Software Co.

CRM system for an insurance client: ETL with Kettle, data validation, testing, and user training material.

Apr – Aug 2020 · Kunming

Operations Intern, Hongta Securities

Collected platform data and reported on operational metrics such as daily and monthly active users.

Yuqing beside a brass astrolabe, books and a balance scale

University of Zurich

Master of Science
in Computer Science

Major in Data Science · Minor in Banking and Finance

2022 – 2025

Yuqing adjusting a brass clockwork mechanism with dividers

Shanghai University

Bachelor of Engineering
in Mechanical Engineering

Mechanical systems · Structural design · Control

2017 – 2021