Android App Safety Engineering (Mar 2024 - present)

SDK risk in Android. Working on LLM model to improve SDK risk detection.

Knowledge Engine Trust & Feedback (Sep 2021 - Mar 2024)

Tooling in Search. Working on a next-gen task management system that leverages user feedback and real-time data streams to identify and address search trust concerns, leading to lasting system-wide enhancements.

Crowdsourcing unlocks solutions computers and humans alone can't achieve efficiently or effectively. But the key challenge lies in ensuring clear communication between requesters and workers. My research tackles this head-on, focusing on task design as the crucial bridge for accurate results and reduced variability.

My dissertation explores how to elevate communication through four key strategies:

1. Targeted incentives: Attract workers with the right skills for the job.

2. Crystal-clear instructions: Leave no room for misinterpretations.

3. Optimal knowledge transfer: Choose the best method to share crucial task information.

4. Rapid AR prototyping: Empower requesters to quickly create immersive, interactive instructions.

Advanced C Programming, Teaching Assistant Fall 2017, Spring 2018, Spring 2019

• Wrote and managed automated testing of programming assignments

• >300 students in Fall 2017; 184 students in Spring 2018; 159 students in Spring 2019

• Responsible for updating course web page and assignment descriptions

• Office hours helping undergraduate students with C programming

■ Built a Chrome Extension to help users easily insert links from multiple sources to webpage

■ Simplified implementation to pure frontend and reduced SLC from 4.2k to 1.3k

■ Implemented UI and data aggregation in TypeScript

■ Used Chrome Extension API to monitor user behavior and insert links to webpage