Hanjie Jiang

Machine Learning Scientist - Product Pricing at Expedia Group

I enjoy solving problems in all fields using modeling approaches and coding languages. I am strongly motivated in researching and analyzing data, and I love learning about new things and exploring the data world. In my spare time, I am bridging the gap between SOTA research and engineering reality through deterministic cognitive pipelines and production-grade ML systems.

Expertise: Pricing ML, Deep Learning, Optimization, Scientific Computing, GenAI, Agentic AI, LLM Systems

The Curator Agent: Bridging the Gap Between SOTA Research and Engineering Reality

The Challenge: The “Hugging Face Daily Papers” Signal-to-Noise Ratio In the rapidly evolving AI landscape, staying updated with SOTA (State-of-the-Art) research is a double-edged sword. Every morning, dozens of papers are released. For a Lead ML Scientist, the manual task of filtering these takes hours, leading to “recommendation system cocoons” or information overload. The Solution: A Modular Cognitive Agent Unlike a linear script, The Curator Agent was built as a Level 1 Agent—a deterministic cognitive pipeline. It doesn’t just “scrape” data; it perceives, thinks, and remembers. ...