Senior Machine Learning Engineer
CurrentLeading Computer Vision projects at KJR.Key responsibilities and projects:• Data and ML Quality Assurance: Contributing and applying our Validation Driven Machine Learning Methodology• Develop and Deploy Models on Edge Devices: Orchestrate the full cycle of developing and deploying object detection models on edge devices. This includes overseeing dataset curation with a focus on differential privacy, establishing rigorous label guides and come up with sampling strategies. I emphasise model training using data-centric approaches, experiment logging, and research in responsible AI. Post-training, I lead efforts in model testing, edge deployment, and active learning strategies to refine and evolve model performance continuously.• MLOps & DevOps Enhancements: Implement MLOps pipelines, enhancing model development lifecycle efficiency. Develop DevOps strategies that optimize software build and deployment processes, ensuring project consistency, speed, and high-quality output.• Geospatial Data Engineering and Remote Sensing: Leading projects utilizing drone data for environmental monitoring, including detecting and counting feral animals with thermal videos and identifying litter hotspots using RGB nadir imagery. I develop optimized capturing strategies for drone pilots, create comprehensive data lakes to manage geospatial data, and work with custom geovideo annotator tools to extract and analyze telemetry data. Additionally, I leverage WebODM for photogrammetry to generate maps and analyze them in QGIS, identifying object counts and hotspots. This end-to-end pipeline encompasses data collection, storage, processing, and analysis, ensuring precise geospatial tagging and actionable insights.