top of page
Google Labs
I worked as a UX Researcher at Google Labs, which is Google’s hub for the latest AI experiments and technologies. I managed end-to-end research workflows for two AI-based consumer applications, focusing on both generative and evaluative research.
Screenshot 2024-04-05 at 11.19.10 AM.png
Oct 2022 - Mar 2024
UX Researcher
AI Test Kitchen
PM - Kristin Yim
Designers - Shirley Leung
Manager - Gabe Clapper

UXR Mentor - Maya Elise, Joyita Banerjee
PM - Raiza Martin
Designers - Jason Spielman
Researchers - Aysha Siddique, Corbin Cunningham, Justin Pacione
Manager - Grace Nicklin
At the onset of the new AI wave marked by the launch of ChatGPT, I witnessed and contributed to the transition of AI becoming central to product development at Google. Specifically, I was involved in building consumer applications, including an AI DJ companion based on the text-to-music model and an AI reading comprehension buddy built on top of Gemini, Google’s generative AI chatbot.
MusicLM: AI DJ Companion
As part of the AI Test Kitchen, which rapidly launches new and innovative AI apps, I assisted the team in monitoring and distilling insights from post-market launch feedback. In addition, I conducted rapid iterative design evaluation studies to inform new iterations of demos. I closely collaborated with the design team during the development of MusicLM, a tool that generates two original music outputs for users to rate, thereby training the model. I led surveys, usability tests and cognitive walkthroughs that identified fail points and areas of friction in the system. I prioritized the identified issues based on frequency and severity. This led to the introduction of an onboarding flow to set the right expectations and appropriate feedback mechanisms to capture Reinforcement Learning from Human Feedback (RLHF) user feedback.

Prompt Input Field (on left) and Audio Player (on right)

NotebookLM: AI Reading Comprehension Buddy 
NotebookLM is a tool designed to enhance how users collect, organize, and interact with project notes and documents through an AI chatbot, introducing retrieval-augmented generation to anchor conversations within users' uploaded documents. Early in its development, I identified a gap in our team's understanding of user needs. To address this, we agreed to undertake foundational research as our initial step. I conducted landscape reviews and task analysis to gain insights into the market and the typical writing workflows, covering the spectrum from initial research that informed our early design decisions to concept testing that confirmed our direction. I also carried out a diary study to understand how users engage with the product over time. These efforts helped shape the product's roadmap and involved generative participatory design with our target users, keeping our development closely aligned with their needs.  

Along the way, we created two compelling artifacts: a user stories deck and a product wrap-up deck that recorded annual achievements and progress, inspired by the Spotify Wrapped presentation. 
Screenshot 2024-04-05 at 11.19.10 AM.png

Source view (on left) and chatbot view (on right) displayed at the same time

These projects are under a non-disclosure agreement.
To know more about my experience ping me at
bottom of page