PFF

About PFF PFF is a world leader in the collection, analysis, and application of sports data. With clients spanning the majority of the professional and college football landscapes, as well as many media entities and other sports fans at large, our employees have the opportunity to drive changes in the way sports are played and consumed by coaches, athletes and consumers. We are looking for a Computer Vision (CV) Engineer/Scientist to join our foundational team and help develop high-performance object detection and tracking models tailored for football sports footage. You will be responsible for deployment of the current architectures within our framework and further improving the accuracy of the models with training on the go. You’ll work closely with a cross-functional team of engineers, analysts, and football experts to push the boundaries of what’s possible in sports tech. What You’ll Do ? Maintain a CI/CD pipeline for CV models and workflows Handle training and fine-tuning of CV models for object detection, tracking, homography, multi-modal analysis, etc., Be proactive in researching on latest CV developments and bring proof of concept projects (POC) Build scalable data engines for evaluation of models and their integration to existing framework Contribute towards development of custom datasets for training and validation Collaborate cross-functionally with data scientists, football analysts, and engineers to deploy CV systems in production Evaluate models with custom dataset Use internal and external APIs including AWS platform Maintain efficient documentation in Git and Confluence Minimum Qualifications 5+ years of experience in Computer Vision or Applied Machine Learning roles. Hands-on experience with modern detection and tracking Strong understanding of projective geometry, camera calibration, and video homography. Proficiency in Python and CV/ML tools such as PyTorch, OpenCV, MMDetection, Ultralytics and relevant platforms. Knowledge of hybrid inference strategies (e.g., cascade models, frame-skipping, multi-stage inference). Experience in deploying models in real-time systems or stream-processing frameworks. Strong Python and CV tooling background: OpenCV, MMDetection, Ultralytics, NumPy, SciPy. Comfort designing evaluation pipelines tailored to real-world use cases and edge cases. Experience with large-scale video datasets and temporal training techniques. Experience in AWS for training and deploying models Preferred Qualifications B.E/B.Tech/M.Sc/M.Tech in relevant field Prior experience with sports video, broadcast, or ball/player tracking. Experience in training and deploying models using AWS The Pay Range For This Role Is 190,000 - 220,000 USD per year (Remote (United States)) 280,000 - 330,000 CAD per year (Remote (Canada))