Quality Assurance internship
Background: I worked as an intern in an AI startup using the Agile Software Development model
Task: Work with the Quality Assurance team to ensure that the firm's products meet industry standards and provide a good user experience
Actions:
- Conducted daily regression tests to identify User Interface (UI) and User Experience (UX) issues
- Used Project-Management Software (ClickUp) to create, maintain, and re-test a backlog of bugs and feature requests
- Reviewed live user session recordings (LogRocket) to discover UX pain-points
- Discussed ideas and prototypes for possible improvements with Project Managers and Development Leads
- Investigated the use of Contrastive Loss functions and in Large Language Model (LLM) development with the Natural Language Processing (NLP) research team
Results:
- Reported and re-tested more than 100 bugs!
- My UI & UX suggestions made it into the final product!
- I used Figma created prototypes for my design ideas and shared them with the developers
- I took responsibility and filled in when members of my team are unavailable
- I participated in daily standup meetings to receive project updates and raise UI/UX concerns
Legal Lingo
Background: My team and I were tasked to address a social problem using deep learning methods.
We drew on our shared experience of translating English documents to our parents, who were
not as familiar with the English language as us
Research Task: (1) Create a translation tool tailored to the high-stakes nature of legal documents
(2) Pitch and present our idea to a audience of judges, educators, and industry leaders
Actions:
- Created Legal Lingo, a web-based translation tool
- Developed the OCR module to scan images of legal documents users uploaded
- Interviewed members of my community who face similar challenges to
understanding legal documents in a foreign language to identify pain-points
- Discussed with project mentors to identify and address limitations and problems
in our project
- Met weekly to design and troubleshoot our product
- Created and practiced our presentation pitch with professional communicators
Results:
- We developed Legal Lingo, a high-impact tool to advance social justice and inequalities
brought by language barriers
- Our project won the 2022's FutureMakers Create-a-thon
- We presented at the 2023 ASU-GSV summit to educators, investors, and industry leaders
Teaching Assitant for Natural Language Processing class
Background: I joined the Natural Language Processing class as a teaching assistant
Task: (1) Grade homework and exams (2) Hold officer hours
Actions:
- Held officer hours to answer questions and provide guidance
- Graded homework and exams using GradeScope
- Created a PyTorch tutorial based on pain-points
I observed from grading and from office hour conversations
Results:
- Every student I helped made significant progress in their homework problems!
- Students made much more progress in their coding tasks and implemented more complex neural network models
Previously, they struggle with the process and structure of training neural networks in PyTorch and spend
much more time than they would like