ResearchBridge
AI Research Platform for equitable research learning and instructor visibility
Role
HCI Researcher
Client
Northwestern Medicine (NM Scholars / Youth Programming)
Timeline
6 months, Spring & Summer 2025
Tools
Figma, GPT-4, Cursor AI
Project Brief
ResearchBridge is a multi-modal research companion for first-generation and underrepresented high school students and their instructors. What began as an AI workbook grew into a two-sided platform: students gain scaffolds that boost confidence, curiosity, and fluency in research; instructors gain visibility and tools to guide engagement equitably at scale.
As HCI researcher, I contributed to the design and evaluation of scaffolded AI modules, instructor dashboards for equitable engagement, and research-backed design decisions. This was about more than AI adoption—it was about designing for students who had been left out of both tech fluency and traditional research pipelines.
Scaling equitable engagement on both sides
For students
Research was overwhelming. Most had limited exposure to scholarly work, AI tools, or independent synthesis, which eroded confidence.
For instructors
With program size doubling, it was nearly impossible to track engagement, spot inequities, or provide timely, individualized support.
The challenge wasn’t just delivering research skills—it was scaling equitable engagement on both sides.
Solution
I contributed to ResearchBridge as a two-sided system:
For Students
- Scaffolded GPT-powered modules for summarizing, questioning, and outlining
- Multi-modal outputs: text, visuals, and reflections
- Built-in ethical check-ins and prompts for critical thinking
- Personalization: everyday vs academic tone, examples tied to student context
For Instructors
A lightweight dashboard surfacing:
- Mode usage patterns (summarize, ask, outline, cite)
- Confidence signals and reflection depth
- Source quality breakdowns (peer-reviewed, gov, news, etc.)
- Alerts for “stuck” students (repeated clarifications, low activity)
Designed as support, not surveillance—focusing on trends and equity, not micromanagement.
Impact
3×
Insight quality
95%
Confidence gains (reported)
100%
Adoption into next-year curriculum
Instructors gained first-ever visibility into research engagement at scale, enabling equitable interventions. The program established a roadmap for NM Scholars to expand AI equity initiatives.
Design choices that shaped the platform
- •Scoped GPT prompts to build confidence, not shortcuts
- •Removed open-ended tasks that caused cognitive overload
- •Integrated reflection checkpoints for ethical reasoning
- •Created instructor-facing summaries that are aggregate, not invasive—balancing student trust with program needs
Reflection
ResearchBridge is more than an AI tool—it’s a platform that rethinks how research can be learned, taught, and scaled equitably. By designing for both students and instructors, the team moved from a one-off workbook experiment to a sustainable system for building research confidence and equity.