News & evidence
Bedtime News Knowledge Atlas
Explore people, places, topics, and events across the Bedtime News archive.
Individual news stories connect to timelines and a knowledge graph, with evidence traced to the original passage.
Research · Software · People
Shiyao Zhang 张世尧
PhD candidate · University of Liverpool
Based at Xi’an Jiaotong-Liverpool University, Suzhou
I study how AI agents can work together to make sense of different systems’ knowledge.
My research brings large language models into multi-agent dialogue and ontology alignment: helping systems agree on what their data means so they can exchange and use it. Alongside research, I build tools for personal AI assistants, agent collaboration, and shared knowledge.
Research
All publicationsHow can agents find common ground?
Two systems may describe the same thing in different ways. I explore how language-model-assisted agents can negotiate these differences and support data exchange across domains.
My work spans ontology alignment, multi-agent dialogue, and resource-centric IoT systems, with publications at AAMAS, ICCCN, and BDDM.
Selected projects
Explore projectsPersonal AI workspace
Ambient Agent
Bring conversations, ongoing tasks, structured knowledge, and interactive apps into a self-hosted personal workspace.
Tasks can pause and resume, while reusable apps work with shared data and capabilities approved by the user.
Reading & attention
Our Choice
Follow the sources you choose in one reader for articles, videos, and podcasts.
A finite inbox and separate discovery page keep subscriptions intentional, while collections and reading preferences stay on your device.
Calendar & privacy
ChronoGraph
Turn selected iOS calendar events into a shareable image while choosing how much detail to reveal.
Select calendars and dates, hide sensitive event details, and generate the image on your device before sharing.
Beyond the code
Making space for shared curiosity.
I started Fun Workshop to bring people across disciplines together for interesting projects, honest insights, and hands-on learning. Through the ClawTwin Initiative, I also brought together more than 20 people to explore how their personal agents could collaborate.
Community & teachingLet’s work together.
I’m interested in research collaborations and opportunities to turn agent research into useful products.
shiyao.zhang14@student.xjtlu.edu.cn