Internship Details

Google Deepmind

At Google DeepMind, I worked with the PAIR (People + AI Research) team to explore how we can make complex AI systems more understandable, trustworthy, and usable—especially for founders building real products. I led mixed-methods research to evaluate the People + AI Guidebook with 30+ founders from around the world, uncovering how trust, transparency, and control shaped their approach to AI adoption.

Student Researcher

3 months

Cambridge, MA

Project Video - Presented at Google I/O

PAIR Guidebook

User Needs + Defining Success

Data + Model Evolution

Mental Models + Expectations

Explainability + Trust

Feedback + Controls

Errors + Graceful Failures

Errors + Graceful Failures

Each week, I co-facilitated sessions to identify real-world use cases and challenges, translating insights into scalable interaction patterns grounded in human-in-the-loop practices. I also collaborated cross-functionally to test internal ML tools—using prompt engineering, RAG, and RLHF to surface and reduce failure cases like hallucinations—and helped present those findings to product and research teams across Google.

More about PAIR

rh692@cornell.edu