About
Hillel Ohayon
B.Sc. in Computer Science and Mathematics, currently completing an M.Sc. in Data Science & AI.
I am an AI Engineer with a strong foundation in data analysis and team leadership. My background includes working with large-scale data systems, machine learning research, and operational analytics. I combine deep technical skills in data engineering (DBT, Airflow) and ML-driven solutions (Python, SQL) with a perspective gained from leading data teams. I enjoy turning complex technical problems into high-impact solutions.
About this project
This website demonstrates the coalition formation algorithm from "AI-Generated Compromises for Coalition Formation" by Eyal Briman, Ehud Shapiro, and Nimrod Talmon (2024).
The algorithm was implemented and this demo was built by Hillel Ohayon for the course "Programming Research Algorithms" at Ariel University.
Algorithm Summary
Given a set of agents with different policy positions and a status quo, the algorithm iteratively proposes AI-generated compromise sentences. Agents vote using a simple rational rule: accept a proposal if it is closer to their ideal position than the current status quo (measured by cosine dissimilarity of sentence embeddings). The process repeats until a majority coalition is formed, guaranteeing that all coalition members genuinely prefer the outcome over the status quo.