Culture & fairness
How can AI-mediated assessment account for different cultural expectations, educational practices, and institutional realities?
RINKY MAHAJAN - HUMAN-CENTERED AI RESEARCHER AND PRODUCT DESIGNER
I study how people use AI.
I build from what I learn.
I investigate how people understand and use AI, especially in education. That research informs my interaction design, alongside hands-on work building AI prototypes.
M.S. in Human–Computer Interaction · Rochester Institute of Technology
3AI projects: 1 published team project, 2 works in progress
5Conference publications: 4 papers, 1 abstract
3+Years industry UX experience
RESEARCH FOCUS
Technology is contextual.
My research starts there.
My work brings human–computer interaction, responsible AI, and educational equity into the same conversation.
Language & learning
How does generative AI shape the ways multiple language learners demonstrate knowledge and engage with academic work?
Transparency & trust
How can human-centered AI design make sources, limitations, and system behavior easier to understand?
SELECTED RESEARCH
The research.
01 / MASTER’S THESIS
Culturally responsive AI assessment
My thesis examines culturally responsive AI assessment with educators, students, and administrators. Research phases are complete; the final draft is in progress.
Mixed methods · Four-country interviewsExplore the thesis →
02 / AI & LEARNING
Language, literacy & generative AI
Understanding how multiple language learners use AI, and how faculty interpret its role in learning.
Interviews · Task-based studyExplore the research →
03 / EXPLAINABLE AI
Transparency in educational AI
A published team project examining course-grounded AI assistance and source transparency.
Human-centered design · RAGExplore the research →
SELECTED PROJECTS
The building.
START HERE / PUBLISHED + EVALUATED
Norman
A course-grounded teaching assistant. How can students judge an AI answer when a source label builds trust but does not let them verify it?
Key finding: Visible sources increased trust without enabling verification.
- My contribution
- UX research, interaction design, usability testing, and paper co-authorship in an RIT team.
- Stage
- Published team project. Initial version evaluated with 16 students; later design responses remain untested.
Sandbox with precomputed answers, not a production deployment.
HI-FI PROTOTYPE / IN PROGRESS
Praxis
How can an at-home rehabilitation session make exercise instructions, camera setup, and recovery from interruptions clear for patients?
- My contribution
- Product and UX design: session flow, wireframes, and safety and accessibility decisions.
- Stage
- Interactive high-fidelity prototype ready, following user flows and wireframes. User testing and technical validation remain next steps.
EARLY UX WORK / IN PROGRESS
E-Val
How can practitioners interpret AI evaluation scores, inspect failed test cases, and compare configurations without treating a score as proof of quality?
- My contribution
- UX and product design: workflow review and framing the comparison experience.
- Stage
- Early interface reviewed; comparison design in progress. No formal user research or usability validation yet.
PUBLICATIONS
From questions to contributions.
Selected conference contributions in responsible AI and inclusive education.
RESEARCH INTO PRACTICE
Investigate how people use AI.
Design from what I learn.
My path brings together industry UX work in public services, healthcare, and finance with applied research at RIT. I connect the practical details of an interaction with the broader questions of who a technology serves.
About my background