Machine Learning Software Engineer
Current• Lead the end-to-end process of computer vision object detection model development, including data preprocessing, feature engineering, training, fine-tuning, and production deployment.• Enhance real-time processing speed by optimizing post-processing workflows for AI object detection on camera feeds, significantly increasing frames per second (FPS) and overall system efficiency.• Create and deploy a comprehensive data communication infrastructure using the Kafka messaging system to facilitate the smooth transmission of flight information and data generated by AI models.• Design and implement an audio classification model using spectrograms and CNNs, managing end-to-end processes from dataset creation to backend integration, enhancing system capabilities.• Overhaul backend service logic for AI model integration, implementing targeted optimizations to reduce database load from ~40 requests per second to 1, enhancing system efficiency and responsiveness.• Implement best practices for software development by introducing version control and establishing robust debugging and testing procedures, including the creation of test beds, enhancing overall system reliability and code quality.• Lead and mentor a team of annotators, providing guidance on meticulous data annotation practices, resulting in high-quality datasets instrumental in designing and training effective object detection models, contributing to the success of AI initiatives.