Jingru (Jessica) Jia
Hello! I am an economist and research scientist at Amazon, where I work on AI agents, causal inference, and experiment calibration. Before joining Amazon, I completed my Ph.D. at the University of Illinois Urbana-Champaign.
I welcome research collaborations.
I mainly focus on the evaluation and post-training of AI agents. I am glad to collaborate in either capacity: through Amazon on applied, industry-facing questions, or independently as an academic researcher on open-ended work. If you are working on something related, please get in touch.
Research
AI agents: evaluation and post-training
I study how AI agents behave: how they reason and make decisions under uncertainty, how they act strategically when other agents are involved, and how post-training shapes that behavior. This work brings the tools of experimental economics to the evaluation of AI systems.
- LLM Strategic Reasoning: Agentic Study through Behavioral Game Theory NeurIPS, 2025
- Decision-Making Behavior Evaluation Framework for LLMs under Uncertain Context NeurIPS, 2024
- An Experimental Study of Competitive Market Behavior Through LLMs NeurIPS 2024 Workshop on Behavioral Machine Learning
- Measuring and Closing the Value Action Gap in LLM AgentsWorking paper
Behavioral and development economics
Earlier in my career, before I became an AI economist, I worked as a behavioral and development economist, designing field experiments and randomized controlled trials in sub-Saharan Africa and Central America.
- Pathways linking WASH access and women's empowerment Journal of Rural Studies, 2025
- Information interventions and willingness to pay for PICS bags Food Policy, 2024
Additional manuscripts from these workstreams are completed and forthcoming.
Contact
The fastest way to reach me is by email, or you can connect with me on LinkedIn.