Projects
CWIReader
Complex Word Identifier and Reader Web App
A web app to enhance reading comprehension by identifying and explaining complex words in texts. I tuned BERT and RoBERTa NLP models to label words based on complexity on the backend.
- Python
- PyTorch
- JavaScript
- React.js
- Jest
- Node.js
- AWS
- Cognito
- Lambda
- SageMaker
- DynamoDB
- Docker
A Nearest-neighbor Approach for Cross-lingual Complex Word Identification
One of the projects I worked on as a Graduate Researcher at WWU. I focused on integrating few-shot learning techniques with LLM fine-tuning to develop a system capable of classifying complex words in multiple languages with minimal training examples. By experimenting with various few-shot algorithms and models, I successfully implemented a solution that significantly improved cross-lingual complex word identification accuracy.
- Python
- PyTorch
- 🤗 Transformers
- WandB
PNWMoths App
Moth Identification and Research Android App
I worked on this mobile application while volunteering for the PNWMoths Project at the WWU Biology Department. I worked as part of an agile team of 4 developers under the mentorship of a former Apple Engineering Director to iterate and improve on features, like cloud storage and geolocation of field data. I made the expert-modifiable key to identify moths out of the 1200+ species in the Pacific Northwest.
- Android SDK
- Android Studio
- Java
- Firebase
- Google Maps SDK
- Git
- JUnit
- Espresso