Computer Vision Engineer

We have an opportunity on the Data Science team at Georgia-Pacific’s Collaboration & Support Center, located in Atlanta, GA. We are looking for an experienced Computer Vision Engineer to join a team of Data Scientists and engineers to focus on providing support and solutions to our manufacturing operations with an emphasis on Computer Vision. Working closely with business partners and engineers, our team supports operations through development and deployment of useful and scalable statistical, machine learning, and deep learning models that make our facilities safe, efficient, and optimal.

What You Will Do

• CV Solution Architecture: Design end-to-end computer vision solutions from the ground up, selecting the right approach across the full spectrum from classical image processing (filtering, morphology, geometric transforms) through to modern deep learning architecture.
• CV Pipeline Ownership: Design, train, and deploy object detection, segmentation, and classification models for defect detection, quality inspection, and worker safety monitoring in active manufacturing environments.
• Project Ownership: Own all aspects of CV projects end-to-end: problem framing, data collection strategy, annotation pipeline setup, preprocessing, model training, validation, deployment, and post-production monitoring.
• Dataset & Annotation Engineering: Build and maintain annotation workflows and dataset versioning practices to support iterative model improvement at scale.
• Production Inference: Develop and manage inference pipelines targeting edge hardware and cloud infrastructure, including latency optimization, throughput tuning, model drift alerting, and own retraining decisions and schedules to maintain model robustness.
• Systems Integration: Integrate CV models into existing manufacturing control systems and IoT sensor streams; collaborate with software and process engineers on system interfaces.
• Partnership: Develop and maintain relationships with key operations and engineering partners; translate physical manufacturing problems into tractable vision problems and communicate findings clearly.
• Pragmatic Delivery: Build toward minimum viable solutions that solve most of the problem; ship working systems that capture value, then iterate. Perfection is reserved for safety-critical applications.

• Bachelor’s Degree or higher in a field related to data science such as Engineering, Mathematics, Business Analytics, Statistics, Computer Science or Data Science
• Proven experience with Computer Vision
• Proven experience with classical CV/image processing using OpenCV: morphological operations, contour analysis, filtering, perspective transformations, and color space manipulation.
• Proven experience with deep learning frameworks (PyTorch or TensorFlow), including hands-on training of CNNs, object detectors, and segmentation models. Must have familiarity with ResNet and YOLO
• Proven years of experience developing production computer vision pipelines: data collection, preprocessing, augmentation, training, and evaluation.
• Proven experience with Git
• Proven experience creatively applying machine learning outside of academic datasets and using a formal programming language (Python or R or SAS)
• Experience in artificial intelligence that encompasses having models deployed to production with successful outcomes for users.

What Will Put You Ahead

• Master’s Degree or higher in a field related to data science such as Data Science, Mathematics, Engineering, Statistics, Computer Science, or Business Analytics
• Experience with containerization and cloud computing (AWS), and a strong understanding of data/compute/memory tradeoffs and optimization for ML workloads
• Experience optimizing models for real-time inference (quantization INT8/FP16, pruning, TensorRT, OpenVINO) and deploying cross-platform runtimes (ONNX Runtime, TensorFlow Lite) on edge targets such as NVIDIA Jetson, Raspberry Pi, or industrial vision controllers.
• Experience with anomaly detection and one class classification for unsupervised defect detection, and with pose estimation or human keypoint detection for ergonomics analysis and safety zone enforcement
• Experience with model explainability tools in a CV context: Grad-CAM, SHAP for image features, or attention visualization
• Experience working with Manufacturing Operations, Consumer Packaged Goods, or IoT data.
• Proven systems-design experience building and operating stable, highly available production deployments for ML/CV systems, including high availability and fault tolerance, blue green rollouts and rollback strategies, CI/CD/model versioning, monitoring, and scaling planning.

You know us already! We make Brawny® paper towels; Dixie® paper cups and plates; Angel Soft® bath tissue; enMotion® paper towel dispensers, DensGlass® gypsum board and Plytanium® plywood you see in your big box home improvement stores and much more! We employ about 35,000 people who want to make a positive difference in today’s world by creating real long-term value for our customers., All Koch companies value diversity of thought, perspectives, aptitudes, experiences, and backgrounds. We are Military Ready and Second Chance employers. Learn more about our hiring philosophy here ., At Koch, employees are empowered to do what they do best to make life better. Learn how our business philosophy helps employees unleash their potential while creating value for themselves and the company.