I’m currently focused on frontend architecture, cloud-native development, DevOps practices, and applied machine learning, with a growing interest in federated learning for healthcare systems. I’m working on a federated learning project for glaucoma detection from retinal fundus images, exploring privacy-preserving model training across simulated hospital clients using the REFUGE dataset and a ResNet-18 model, with the goal of enabling secure collaboration between healthcare institutions without centralizing sensitive patient data.

The project compares centralized training against FedAvg and FedProx, specifically investigating how uneven data distributions across hospitals affect model performance, and evaluates secure aggregation and differential privacy as additional protections against leaking patient information through model updates.

My work combines modern web interfaces with cloud infrastructure to support model training, deployment, and monitoring. I’m especially interested in building practical ML pipelines, improving system reliability, and understanding how frontend applications integrate with backend services and distributed learning environments.

I enjoy learning through hands-on projects, refining my engineering skills, and applying what I learn to real-world healthcare and data-driven use cases. I’m always open to meaningful collaborations and opportunities where I can build, learn, and contribute value.