FEATURED WORK
Careergram
An AI-augmented career exploration platform that aims to spur new insight and decision-making for youth programs to support youths’ occupational identities.
Integrated OpenAI's ChatGPT to enable personalized reflection, leveraging AI-powered natural language processing. Converted existing REST API to GraphQL for database management, streamlining data retrieval processes.

Edsight
A data analytics platform designed to assist educators in improving their teaching pedagogy through data and reflection. It is part of a broader initiative called PMR2 to develop a system to support the implementation of instructional improvement strategies in middle-school mathematics.
Built the course settings page, implementing database interactions with CRUD operations and improving the usability of course management features.

GAN Augmentation
Analyzing spaceflight radiation is challenging due to limited available data. We addressed this by augmenting emperical data with synthetic images.
Trained a Generative Adversarial Network (GAN) to generate images for NASA dataset hosted on AWS. Assessed the efficacy of the GAN-generated images with a ResNet model, achieving a 78% validation accuracy. Optimized model performance through hyperparameter tuning with sweeps and benchmarked results.

Happy Home
A minimalist all-in-one food, water, fitness, and period tracker to aid you in keeping track of your well-being.
In 36 hours, conceptualized, designed, and developed a user-first well-being tracker. This mobile application offers seamless, real-time synchronization with a cloud database, complemented by user authentication to ensure data security.

Directed Acyclic Graphs
The Human Systems Risk Board (HSRB) at NASA created directed acyclic graphs (DAGs) to communicate the causal flow of risk from spaceflight environmental hazards to health outcomes.
Validated expert-drawn DAGs by generating probabilistic graphical models, populated with empirical biological data from the NASA Open Science Data Repository (OSDR). Researched the NoDAG algorithm to support the HSRB in formalizing a shared causal flow model for astronaut medical risks assessment.
