At the Feminine Intelligence Agency (FIA), our mission is to provide women with data-backed insight into relationship dynamics, especially in the early stages — when awareness can still lead to prevention. This project invites students to work on a groundbreaking AI tool that helps identify patterns of emotional manipulation and predict trauma outcomes before they escalate into serious psychological harm.

Using structured synthetic datasets generated by our Player Identifier Chatbot, students will help us explore the difference between benign behaviors and early-stage coercion tactics — building visualizations, dashboards, and models that make emotional risk understandable and actionable for everyday users.

The goal is to create visual tools that empower women to distinguish between what’s emotionally immature and what’s actually dangerous — reducing false positives, increasing early detection, and improving digital safety for people navigating high-stakes interpersonal dynamics.

Project Description / Deliverables:

This project invites students to collaborate with the Feminine Intelligence Agency (FIA) on the development of our Player Identifier Chatbot — an AI tool designed to detect early-stage patterns of psychological manipulation in relationships. Our mission is to help women understand emotional risk before serious harm occurs, using data to distinguish between annoying behavior, emotional immaturity, and patterns that signal trauma-inducing abuse.

While we do not have 6 months of historical user data, we will provide a structured, high-quality synthetic dataset, generated through a finely tuned scoring system based on psychological and behavioral research. This ensures students can work with complex data while avoiding privacy concerns.

Your Goal:

Design visualizations and data models that help users see and understand their emotional risk clearly and empathetically — empowering better decision-making in high-stakes relational contexts.

🧪 Potential Deliverables:

• A Risk Profile Dashboard that breaks down red flags, behavioral traits, and predicted trauma outcomes
• A “Lock-and-Key” visual map illustrating how specific abuser types target particular cognitive or emotional vulnerabilities
• A timeline or heatmap showing when emotional abuse escalates or becomes more detectable
• Comparative visuals showing how high-intensity (but non-harmful) neurodivergent behavior differs from manipulative patterns
• (Optional) Classification models or clustering algorithms to explore how well the AI identifies specific player types

🛠️ Tools & Techniques (optional, depending on student interest):

• Python / Jupyter Notebooks
• Tableau or Power BI
• Logistic regression, decision trees, or unsupervised clustering
• Plotly, Seaborn, or D3.js for interactive visuals
• UX design thinking for trauma-informed visual communication

This project is ideal for students interested in:

• Data science for social good and gender equity
• Ethical AI and behavior modeling
• Visualization of abstract concepts like emotional risk, manipulation, and trauma prediction
• Building tools that make a real-world impact

We’ll provide:

• A rich, structured synthetic dataset
• A scoring system and profile examples
• Feedback and support on design thinking and mission alignment