Case Study

A research-driven AI solution for adolescent health

Summary

Research leadership and AI platform integration • May-July 2025

Led systematic AI evaluation and UX research for educational chatbot serving adolescents aged 8-14 with menstrual health content. Applied responsible AI development practices to identify critical market gaps, delivering strategic positioning that achieved exceptional client satisfaction and prototype quality indistinguishable from production applications.

Challenge

Adolescents aged 8-14 lack access to stigma-free, age-appropriate menstrual health education. Existing AI solutions either target older teens (13+) or provide medically accurate but emotionally sterile responses that fail young users.

My Role

  • UX Research Lead and Initial Project Manager

  • Led 3-person UX team coordinating with 4-person development team

  • Conducted systematic competitive analysis across AI platforms

  • Delivered client presentations and strategic recommendations

Research Approach

Competitive Analysis Methodology

I evaluated AI platforms using standardized educational scenarios to assess age-appropriateness, emotional intelligence, and cultural sensitivity.

Key Testing Questions:

  • "What is menstruation?" (age-appropriate explanations)

  • "I'm at school and just got my period" (crisis response)

  • Religious and cultural scenarios (inclusive design)

Competitors Analyzed

Planned Parenthood's "Roo":

  • Medically accurate but emotionally sterile responses

  • Targets teens 13+, leaving gap in 8-12 age range

  • Rigid conversation flows lacking empathy

OpenAI’s ChatGPT and similar AI Platforms:

  • Highly adaptable and empathetic responses

  • No built-in age controls or educational safety measures

  • Privacy concerns for vulnerable populations

Research Deliverables

  • Competitive analysis report (12 pages)

  • Research presentation slides (20-minute client presentation)

  • Strategic positioning recommendations framework

  • AI platform evaluation documentation

Key Research Insights

Market Gap Discovery

Identified critical underserved population: No existing solutions effectively serve elementary/middle school students (ages 8-12) with appropriate tone and emotional support. Despite menstruation beginning, on average, when a child turns 12, existing products primarily focused on those 13 and older.

User Behavior Finding

Young users often don't know what questions to ask about menstruation, requiring conversation starters and guided discovery rather than reactive Q&A.

Cultural Sensitivity Gaps

AI-powered platforms for puberty education inadequately handled religious and cultural considerations, missing opportunities for inclusive health education.

Strategic Recommendations

AI Development Priorities

  • Built-in age-responsive design with automatic tone adjustment

  • Panic recognition and comfort provision beyond clinical information

  • Privacy-first architecture with ephemeral conversations

Cultural Competency Framework

  • Respectful handling of diverse family structures and religious practices

  • Myth-busting approach with cultural sensitivity

  • Inclusive language supporting all backgrounds and personal identities

Design Process

Wireframing and Prototyping

In addition to creating some early sketches, I created user flow documentation and provided extensive feedback on wireframes developed by a UX designer on our team. We progressed from initial concepts to high-fidelity prototypes with a comprehensive style guide for handoff to the engineering team.

Research

The wireframing and prototyping work proceeded on a parallel track with competitive research, allowing us to meet compressed client timeline requirements while building strategic foundation for future iterations.

Conversation Design

I developed sample conversation examples using ChatGPT with an uploaded knowledge base of vetted resources (50+ HTML files) to demonstrate realistic AI response patterns. This testing informed technical constraints and provided the engineering team with additional considerations.

Additional Challenges

Limited Access to Target Users

Direct user testing with 8-14 year olds would required parental consent and specialized protocols we couldn't implement within timeline constraints. I addressed this recruiting an additional UX researcher with a pre-teen family member interested in providing feedback and testing.

Technical AI Constraints

Balancing educational UX ideals with the OpenAI platform limitations and $20 budget required creative workarounds. I focused on conversation design examples that demonstrated potential within technical feasibility.

Results and Impact

Client Success

  • Delivered 20-minute research presentation achieving exceptional client satisfaction

  • Created prototype quality indistinguishable from production application

  • Research insights directly informed client's competitive positioning strategy

Strategic Value

  • Identified clear market opportunity in underserved age demographic

  • Provided framework for responsible AI implementation in education

  • Established competitive differentiation against established players

Core Contributions

Research Innovation in New Domain

Successfully adapted systematic competitive analysis methodology to AI platform evaluation, identifying critical market gaps in educational technology while building expertise in responsible AI practices for vulnerable populations.

Cross-Functional Project Coordination

Successfully managed research, design, and client stakeholder alignment within compressed timeline while maintaining quality standards and ensuring seamless deliverable handoff.

Tools and Methods

  • Research: Competitive analysis, AI platform evaluation, user scenario testing

  • Prototyping: Figma for wireframes and prototypes, style guide development

  • Project Management: Notion and Otter for documentation, Google Drive for asset management

  • AI Integration: OpenAI platform assessment, conversation design, technical constraint evaluation

Want to see more case studies?

Click or tap here.