Curriculum Associates

Leveraging the latest Generative AI technologies to improve the company's business and product development processes.

Principal Data Scientist / AI Engineer

Standard AI

Led ML algorithm development for automatic item recognition in streaming video data, transforming shopping experiences for convenience stores and small markets. Developed an XGBoost classifier model working in conjunction with a deep learning image change detection model to identify purchased items by recognizing interactions with store shelves and determining clear view times. Deployed the new algorithm to all stores, integrating outputs into PubSub messages for system-wide utilization. Planned the development of a complex deep learning multi-view/multi-modal model, synchronizing inputs from video clips, 3D pose tracking, and shelf images to more accurately determine items.

Sped up image preprocessing by 80% utilizing Torchscript on Google Cloud VertexAI.  Developed method to correct for camera motions and identify failures in the occlusion algorithm, leading to a reduction in false positives. Created Python/Plotly store monitoring and model performance dashboards, improving algorithm behavior and facilitating thorough verification.

Staff Machine Learning Engineer

Leidos

Led a team of eight engineers and data scientists to develop real-time computer vision and machine learning software for detecting explosives and weapons in 2D/3D X-ray baggage images. Designed high-level solutions, developed core software, and managed the full algorithm development life-cycle, ensuring efficient and effective implementation. Created a C++/Python/R algorithm software framework for rapid development and seamless deployment of detection algorithms on real-time scanners. Applied machine learning and deep learning (CNNs) to enhance detection performance and reduce training data requirements.

Projects involved agile development of 2D/3D image segmentation, object detection, feature extraction, machine learning, deep learning, data science, data collection, and product-quality software development. 

Collaborated with product managers to prioritize deliverables and estimate project costs and schedules to ensure timely and cost-effective project completion. Defined performance metrics and generated reports to track project progress, enabling effective monitoring and decision-making. Interacted with executives, customers, and government agencies, providing regular updates and managing regulatory submissions.

Director of Computer Vision and Machine Learning

Textron Defense Systems

Developed real-time model-based computer vision algorithms to classify vehicles from infrared and synthetic-aperture radar (SAR) images. Researched and developed image understanding, symbolic matching and pattern recognition algorithms. Created interactive multimedia simulation environment to support testing and demonstrations.

Senior Software Engineer

MIT Lincoln Laboratory

Researched and developed automatic classification and feature extraction algorithms for the real-time prediction and detection of wind shear from Doppler weather radar image using machine learning, time series forecasting, pattern recognition and image processing techniques. Algorithms served as the basis for a nationwide system of wind shear detection radars (TDWR) in use at major airports.

Associate Staff

IBM Watson Research Lab

Worked in the Speech Recognition and Machine Translation Group. Developed machine learning software to automatically annotate French sentences with part-of-speech labels. A large corpus of unannotated French sentences were used to train a Hidden-Markov model; the annotations produced were 97% correct

Research Assistant

Highly-analytical Algorithm Development Engineer/Manager with over 20 years experience in Machine Learning, Computer Vision, Artificial Intelligence, Image Processing/Segmentation, Software Development, and Engineering Management. Strong engineering technical leader with an M.S. in Computer Science and Engineering from the University of Michigan.

Expert in researching, evaluating, implementing, and extending AI/machine learning methodologies to develop deployable software across diverse domains. Adept at providing technical leadership and empowering teams to overcome challenges and deliver superior solutions in complex environments. Equipped with strong problem-solving and collaboration skills: history of steering successful resolution of technical challenges. Instrumental in the growth and success of organizations, notably contributing to a startup's algorithmic foundation that propelled a successful IPO and transformation into an industry leader in the explosives detection domain.

Taplio

Jeff Stillson's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Jeff Stillson's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Jeff Stillson's Linkedin Analytics