Global Senate Policy Researcher

Project Overview

The Global Senate Policy Researcher is an interactive tool designed to break through "Western-centric" or "local-only" policy biases. It deploys 10 specialized AI agents, each programmed to identify as a researcher from a distinct global region (e.g., East Asia, Sub-Saharan Africa, Latin America, Nordic Countries).

When given a policy problem, these agents simultaneously scour their "regions" for best practices, historical precedents, and unique cultural solutions. The system then aggregates these findings into a "Comparative Global Report," offering a synthesized world-view of solutions that a single human researcher might miss.

How It Works

The application uses a "Mixture of Experts" approach, leveraging Large Language Models (LLMs) via the Hugging Face Inference API. It orchestrates a parallel workflow:

1. User Input: You enter a "wicked problem" (e.g., "Combating urban heat islands" or "Regulating gig economy labor").
2. Regional Dispatch: The system sends this prompt to 10 separate agent personas.
3. Local Search: Each agent applies a region-specific lens to find concrete examples and case studies.
4. Synthesis: A final "Editor" agent compiles the results, highlighting conflicts, consensus, and innovative outliers.

Usage Instructions

To use the tool effectively:

1. Click the "Launch" button above to open the Hugging Face Space.
2. Enter your query in the text box. Be specific for better results (e.g., instead of "Education," try "Vocational training programs for at-risk youth").
3. Wait for the agents. It may take 30-60 seconds for all 10 agents to report back.
4. Review the Report. The final output will be structured by themes, comparing how different parts of the world approach your problem.