Role: Team Lead
The goal for the project is to democratize access to resources by developing a freely accessible webapp for identifying submitted chest X-ray images for the following respiratory lung disorders: tuberculosis, lung cancer, pneumonia, and COVID.
Four teams worked alongside each other in building models for each disease for 8 weeks, and within each team, we selected the best model for deployment.
I led the Tuberculosis team through the project tasks: data collection, EDA and data preprocessing, model building and evaluation, and model deployment.
Links:
- Official Project Repo: https://github.com/OmdenaAI/myanmar-chapter-chest-x-rays
- My Project Repo: https://tinyurl.com/Tuberculosis-Detector
- Streamlit App: https://chest-xrays-detection-system.streamlit.app/
Role: Lead Machine Learning Engineer
We collaborated in this OmdenaLore AI challenge with the Giga team, a joint initiative between UNICEF and ITU for two months. We built several Computer Vision and Deep Learning models to detect school locations in Sudan using Satellite Imagery. We did an extensive and thorough analysis of the data and built multiple models using datasets provided by the Giga team to solve this problem.
Role: Data Research Analyst
Attended DataDive hackathon event and participated in continued efforts in data wrangling and visualization for project sponsored by CDAC at UChicago to develop new tools to measure broadband access in US