Responsibilities include:

• Conducting research on foundational models for anomaly detection and segmentation in medical imaging, with a focus on ultrasound data.
• Collaborating with healthcare professionals to design tools that support diagnostic decisions during both in-person and remote examinations.
• Engineered a real-time system integrating object detection and segmentation algorithms to provide insights during medical assessments.

Responsibilities include:

• Conducted state-of-the-art analysis and explored new research approaches for anomaly detection in ultrasound imaging, addressing challenges related to data scarcity.
• Designed and implemented a novel Test-Time Training (TTT) algorithm, published at WACV 2025.

Responsibilities include:

• Analyzed the problem to define requirement, then designed and implemented a data pipeline and model deployment strategy through an iterative approach.
• Implemented machine learning models for document classification, orientation detection, and content extraction (e.g.,names, medical billing details)
• Enhanced model reliability by integrating rule‐based corrections and error‐flagging mechanisms to refine predictions.