Machine Learning Engineer

Syngenta is an agriculture technology company developing seed innovations and digital solutions to support farmers and advance sustainable agriculture. The Machine Learning Engineer will develop and deploy production-grade computer vision and machine learning solutions using imagery and sensor data to accelerate plant breeding programs. The role also involves building scalable data pipelines, cloud infrastructure, and AI products while collaborating with researchers, breeders, product managers, and engineering teams.

Responsibilities:

• Design, develop, and deploy production-grade computer vision models that extract quantitative digital traits from multi-modal imagery (e.g., RGB, multispectral, thermal, hyperspectral, LiDAR, 3D point clouds) captured from drones, ground-based platforms, mobile devices, satellites and other kinds of sensors
• Build and maintain scalable phenomics pipelines that process thousands of field plots across multiple breeding programs, integrating image acquisition, preprocessing, trait extraction, quality control, and delivery to downstream data products with minimal manual intervention
• Collaborate with plant breeders, researchers, product managers, engineers, and data scientists to translate objectives into computer vision and machine learning solutions, validate outputs against ground truth, and ensure scientific and business relevance
• Shape the strategic direction for computer vision in phenomics, defining how to maximize value from proprietary imagery and sensor data through modern ML approaches (self-supervised learning, multi-modal fusion) while balancing innovation with practical deployment needs
• Contribute across the full lifecycle of machine learning projects, including problem definition, data exploration, model selection, performance evaluation, deployment, and monitoring, which could include both phenomics and broader AI/ML applications
• Design, build, and own cloud-based data pipelines and workflow orchestrators to ingest, validate, transform, and deliver imagery and sensor-derived features at scale
• Drive productionalization of research code into maintainable services and pipelines, and optimize existing machine learning systems for performance, scalability, and reliability by applying best practices in software engineering, MLOps/CI-CD, containerization, infrastructure-as-code, and cloud deployment
• Architect and deploy mobile-first AI products that enable breeders to capture images and receive real-time identification, classification, or trait measurements
• Develop and operate automated image preprocessing and quality-control workflows to reliably transform raw imagery into analysis-ready data
• Contribute to knowledge sharing, documentation, and team learning, communicating complex machine learning concepts to non-technical stakeholders and supporting the team's knowledge base
• Follow an agile way of working and collaborating effectively across disciplines and global teams

Requirements:

• Candidates must reside in and be permanently authorized to work in the United States without current or future employer sponsorship
• 5+ years of experience in machine learning engineering and data science roles with 4+ years in applied computer vision
• Deep expertise in deep learning architectures for computer vision (CNNs, vision transformers, segmentation and detection models, etc.) and experience with machine learning frameworks (PyTorch, TensorFlow, Keras, scikit-learn, XGBoost) applied to both imagery and other modalities
• Demonstrated ability to productionalize ML models using strong Python and SQL engineering practices (packaging, testing, code review, Git), MLOps tooling (e.g., MLflow, Weights & Biases), containerization (Docker), CI/CD, and one or more cloud platforms (AWS, GCP, Azure)
• Solid understanding of data structures, algorithms, statistical methods, and workflow management tools for end-to-end modeling, calibration, validation, and application
• Hands-on experience with data engineering and orchestration patterns (ETL/ELT, batch vs. streaming, backfills, idempotency), building and operating ML and data pipelines using workflow orchestrators (e.g., Airflow/Argo/Kubeflow/Prefect) and cloud-native services (e.g., object storage, managed compute, message queues, data warehouses)
• Domain knowledge related to the development and deploying computer vision models specifically for plant phenotyping, agricultural applications, or biological imaging in research or commercial environments
• Knowledge of self-supervised learning, foundation models, transfer learning, and active learning approaches for building generalizable representations
• Master's or Doctoral degree in Computer Science, Remote Sensing, Engineering, Mathematics/Statistics, Geosciences or a related technical field with strong foundations in geospatial analysis, image processing, and machine learning is highly desirable