The Feminine Intelligence Agency (FIA) is a research and prevention lab dedicated to understanding and disrupting manipulation and coercive control—especially as they affect women, neurodivergent individuals, and survivors of trauma. We develop cutting-edge tools that help users recognize patterns of psychological abuse, from relationship fraud to workplace coercion.

Our work integrates behavioral science, survivor testimony, and psychological trait modeling into a unique research framework that makes invisible forms of harm measurable and actionable. We are building tools such as:

• A Player Identifier Chatbot that detects manipulative behavior patterns,
• Simulation environments to teach resistance to psychological pressure,
• A trait-based scoring system (UNES) used to encode manipulative personalities and predict trauma risk,
• And a peer-to-peer support system for skill-based emotional resilience.

We operate like a field lab—constantly collecting case data, building synthetic profiles, and analyzing social dynamics. But we're also increasingly facing a scale problem: the scope of coercion is vast, and the manual curation of insights is too slow to meet demand. That’s where AI agents come in.

This project is both an analytical deep dive and a creative research mission. The student will explore the current state of AI agents (e.g., LangChain, CrewAI, Autogen) and map out how these tools could power FIA’s next-generation platform. They’ll analyze our internal datasets, knowledge structures, and user-facing needs to propose a bold, technically grounded vision for how we can use AI agents to:

• Automate social trait tagging across case libraries
• Scaffold hypothesis testing in trauma and abuse research
• Power chat-based learning systems for survivor discernment
• Coordinate decentralized research across global contributors
• Generate structured outputs from chaotic relational data
• Track patterns in coercive control and relationship fraud
• Create semi-autonomous research assistants with specific “expertise”

This is not a build project—it’s an architectural research and design project. The student will research what’s possible, what’s emerging, and what hasn’t been done yet.

Deliverables:

• Deep landscape review of the current AI agent ecosystem (tools, use cases, risks)
• Audit of FIA’s unique data assets and technical infrastructure
• Vision paper outlining a bold, modular roadmap for AI agent integration
• Design proposals for 3–5 high-value agent roles FIA could implement
• Stakeholder-ready presentation summarizing insights, priorities, and next steps