Skip to main content

CV

Haoqing Wang's CV: education, research experience, publications, open-source and community work, skills, awards, and activities.

Education

Beijing University of Posts and Telecommunications (BUPT) & Queen Mary University of London (QMUL), B.Eng. in Internet of Things Engineering (Joint Programme), Sep 2024–present (expected 2028)
  • GPA: 3.9/4.0

Beijing No. 12 High School, Beijing, China, Sep 2021–Jun 2024

Research Experience

Tencent Hunyuan, Research Intern, Foundation Model Department, TEG, Beijing, China
Nov 2026–present
  • Research: AI4AI
Microsoft Research Asia (MSRA), Research Intern, Intelligent Multimedia Group, Beijing, China
Mar 2026–Nov 2026
  • Mentored by Xiaoyi Zhang
  • Research: Computer Use Agent (CUA) and GUI Agent
Tsinghua University Knowledge Engineering Group (THUKEG), Research Intern, Beijing, China
Nov 2025–Mar 2026
  • Mentored by Zijun Yao and Ji Qi
  • Research: multimodal large language models (MLLMs), video understanding, and long-video token compression
Beijing University of Posts and Telecommunications (BUPT), Research Intern, School of Cyberspace Security, Beijing, China
Jun 2025–Nov 2025
  • Mentored by Jiawen Qian and Hang Du
  • Research: multimodal online video understanding

Publications

  • Hybrid Computer Use: A Unified Programmatic, Semantic, and Visual Interface for General Agents. First author. arXiv preprint arXiv:2610.41432, 2026. [Paper] [Project] [Code]
  • Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use from Evolved Harnesses. Co-first author. Technical report, 2026. [Paper] [Project] [Code] [Model]
  • WebHarbor: Docking Real Websites for Evolving GUI Agent Environments. Research work, 2026. [Project] [Code]
  • Mage-VL: An Efficient Codec-Native Streaming Multimodal Foundation Model. arXiv preprint arXiv:2607.24904, 2026. [Paper] [Project] [Code]
  • Learning to Respond: A Large-Scale Benchmark and Progressive Learning Framework for Trigger-Centric Online Video Understanding. Under review, 2025.

Open Source & Community

Apache Software Foundation, Committer
Aug 2026–present

Apache Maka (Incubating), Initial Committer & PPMC Member
Aug 2026–present

  • Contribute to automated tasks and Computer Use capabilities in the Apache Maka agent workspace.

Open CoDesign (OpenCoworkAI), Project Lead
Apr 2026–present

  • Open-source Claude Design alternative that turns prompts into prototypes, slide decks, and marketing assets locally.
  • Led the product direction and the core agentic design architecture: a design-specific tool loop, starter artifact workflow, syntax and visual verification harness, local snapshot history, and multi-model BYOK support.
  • Reached 2k GitHub stars within 3 days of launch.

hapi, Collaborator
Apr 2026–present

  • Invited by weishu to contribute to core features of a mobile AI coding app that supports Claude Code, Codex, Gemini, OpenCode, and more.

open-cowork (OpenCoworkAI), Core Contributor
Jan 2026–present

  • Open-source Claude Cowork desktop app for Windows and macOS; contributed to feature development, product iteration, and community collaboration.

Personal Projects

  • skill-optimizer (creator and maintainer, 2026–present): diagnoses and optimizes Agent Skill files from real session transcripts and static analysis.
  • AI Course Notes (creator and maintainer, 2025–present): 220+ Chinese PDF lecture notes for AI/LLM courses, generated with a Whisper transcription and LaTeX pipeline.

Skills

  • Programming: Python
  • Machine Learning: PyTorch
  • Tools: Git, Linux
  • Writing: LaTeX
  • Language: English (CET-6: 549)

Awards & Honors

  • National Scholarship (Top 0.2%) (Oct 2025)
  • Outstanding Student (University-level) (Oct 2025)

Leadership & Activities

  • Class Monitor
  • Member, Track and Field Team
  • Member, Football (Soccer) Team