Carey Okal
I'm an MPS student in Information Science at Cornell Bowers CIS (B.S. Information Science '26, cum laude), working at the intersection of LLMs, social simulation, and data science. At the Social Dynamics Lab I audit multi-agent LLM simulations and built the large-scale ingestion pipelines behind persuasion-strategy research on political news; at CIDER I'm co-authoring a white paper on predictive systems for disaster resilience. Before Cornell, I worked with EducationUSA in Nairobi, reaching 20,000 students across 30 Kenyan high schools.
- Social Dynamics Lab · Jun 2026 – Present · Ongoing
AI Social Simulation: LLM Agents in a Virtual World
A multi-agent LLM simulation inspired by the Stanford Prison Experiment. Autonomous agents inhabit a virtual world, form relationships, and interact in real time, and audit pipelines verify what they say against the world state.
Read more → - Social Dynamics Lab · Feb – Jun 2026 · Concluded
Persuasion Strategy Classification in Political News
The ingestion and cleaning pipelines behind a roughly 2-million-article corpus of U.S. political news (14 outlets, 2000–2025), built for NSF-funded research on how persuasive style varies with editorial credibility and ideology.
Read more → - CIDER · Cornell · Mar 2026 – Present · White paper in progress
Predictive Systems for Disaster Resilience
Co-authoring a white paper for Cornell's "Data Science to Build Resilience and Improve Humanitarian Response" Thought Summit, synthesizing submitted research on fusing CDR, satellite imagery, and survey data into early-warning systems.
Read more → - SMART Fellowship · Cornell · Oct 2025 – Apr 2026 · Hackathon winner
Sustainable Ecotourism Commerce Platform
Led the hackathon-winning Community Tourism Relay for Il Ngwesi Conservancy, a booking platform, handmade-goods marketplace, and community revenue portal judged the top solution among 5 multi-institutional teams.
Read more → View the live platform ↗
Research aside, I build for the web: client work and my own products, all currently live.
Open to research collaborations, PhD conversations, and hard data problems.
Email me →