Experience

• 

  Athens information technology

  Jun 2019 - Jul 2019

  Undergraduate research fellow
• 

  Quantiphi

  Jul 2020 - Jun 2022

  Machine learning engineer

  • Delivered a healthcare claims processing solution handling 1M+ claims monthly, reducing need for manual review from 40% to 1% by training machine learning models to extract key information from documents.• Optimized SQL queries and database performance to streamline data retrieval and processing for medical claims, improving overall system efficiency and data accuracy.• Deployed YOLO object detection on AWS Sagemaker to detect forklifts from CCTV footage, improvingallocation efficiency by 30% and generating cost savings for an automotive OEM.• Improved accuracy in extracting critical data from ophthalmic reports by 20% through leading theexperimentation and deployment of ML models for an eyewear company project.• Enabled customers to upload prescription photos instead of manually entering details, enhancing user experience and reducing data entry time. Show less

• 

  Gilead sciences

  Jan 2023 - Jun 2023

  Advanced analytics engineer

  • Increased patient retention by 15% by developing a two-stage XGBoost model using structured and unstructured clinical data to identify patients likely to discontinue medication, ensuring GDPR compliance.• Developed an end-to-end ETL pipeline using Azure Data Factory to ingest and process clinical data from EPIC, improving data processing efficiency by 20% and scaling ML model deployment.• Created interactive dashboards with data visualizations to provide actionable insights.

• 

  Valeo

  Sept 2023 - Jun 2024

  Machine learning software engineer

  • Automated data pipelines to enhance processing and analysis of map data for Advanced Driver Assistance Systems, resulting in a 30% increase in efficiency. Presented key findings at CES 2024.• Led data collection efforts and managed a team to label data, integrating a segmentation model with annotation tools that achieved a 4x increase in annotation speed, crucial for ADAS development.• Trained a machine learning semantic segmentation model, achieving an 8% improvement in accuracyfor parking environments, which significantly enhanced Advanced Driver Assistance Systems. Show less