AI Product Manager building context-aware products at AirJelly
Agent Context Β· Memory Β· Proactive AI Β· Product Systems
- π§βπ» AI Product Manager Intern at Low Entropy AI Β· AirJelly in Beijing
- πΌ Previously an AI Research Intern at Emotional Byte AI Inc., a U.S. AI startup, researching Agent interaction patterns and delivering a bilingual social product
- π B.Mgmt. in Marketing at Shenzhen Technology University, Class of 2027
- π§ Interested in how AI can understand people over time and turn Context into useful action
- π§ Working across product architecture, interaction design, research, and product-engineering collaboration
- π£οΈ Native Cantonese speaker, former debate-team captain, and usually surrounded by cats
At AirJelly, I'm helping shape a product system around Agent Context and context consumption: how scattered user Context is composed, ranked, controlled, recalled, and made useful to both users and external Agents.
My focus is not adding more AI features. It is making AI assistance feel more continuous, timely, understandable, and controllable.
The Reinforcement of the Interaction between High-Speed Rail and Digital Infrastructure on Digital Inclusion-Mental Health: Evidence from China
Paper accepted and presented at the 20th Annual Conference of the International Association for China Planning (IACP), July 2026.
AI Applications Will Survive β Just Not in Their Current Form
Models will keep absorbing today's features, but infrastructure will not invent the products of the AI era by itself.
Two Flywheels: A Management Model Built for AI Startups
A structure I encountered during my first internship β and have continued thinking about since.


