Emmanuel – Python, Tensorflow, Machine learning

The project is a cutting-edge innovator in retail management solutions. It showcases a commitment to leveraging advanced technology to drive operational efficiency and elevate customer satisfaction. A state-of-the-art People Counter system that harnesses the power of store surveillance cameras and Computer Vision algorithms was engineered. With precision, it tracks and tallies incoming and outgoing foot traffic in real-time, offering invaluable insights into customer flow. This innovative implementation empowers businesses to allocate staff effectively and optimize customer service. The People Counter project is a testament to the transformative potential of AI-powered solutions in retail management, redefining how businesses enhance their operations and customer experiences.

• Led and developed a multithreading system for inferences in real time;
• Engineered a sophisticated People Counter system utilizing store surveillance cameras;
• Provided real-time data on customer flow;
• Utilized store surveillance cameras for accurate tracking of foot traffic;
• Improved customer satisfaction through data-driven insights.

The system is a game-changer for retailers, as it continuously monitors and identifies excessive queue lengths in real time. By harnessing the power of cutting-edge technology, including advanced video analysis, our system accurately assesses both customer wait times and queue length, ensuring a proactive response to any issues. One of the key features of our Crowded Queue Alert System is its ability to promptly issue alerts when predefined thresholds for crowded queues are exceeded. This real-time notification mechanism empowers retailers to take immediate action, optimizing staff allocation and significantly reducing customer wait times.

• Designed and implemented the system for real-time alerts on crowded cashier queues;
• Utilized cutting-edge technology to monitor and analyze video feeds from checkout areas;
• Accurately assessed customer wait times and queue length through the system;
• Promptly issued alerts when predefined thresholds for excessive queue lengths were exceeded;
• Improved customer satisfaction by minimizing checkout wait times;
• Optimized staff allocation based on queue data and alerts;
• Demonstrated the potential of AI-driven strategies in retail management;
• Showcased a proactive approach to enhancing customer experiences and operational efficiency.

An innovative Autonomous Robotic System designed for efficient and accurate shelf auditing in a retail environment, harnessing the power of AI and Computer Vision to perform real-time shelf analysis. This innovation not only optimizes inventory management but also drastically reduces the potential for human error, ensuring the highest level of accuracy in retail audits. The team integrated cutting-edge hardware components with advanced software modules to create a user-friendly solution that is as efficient as it is effective. The deployment led to significant time and cost savings, revolutionizing the auditing process and enhancing store operations.

• Trained object detection model;
• Developed the inference pipeline;
• Lead Hardware Communication using ROS
• Led the development of the project;
• Archived the Robotic System for Shelf Audit working for 1 year straight already;
• Reduced obsolete price tags on the shelves from ~16.7% to ~3.4% daily.

The company specializes in leveraging advanced technology to transform the retail industry. Emmanuel pioneered a groundbreaking initiative to create Heat Maps using store security cameras, harnessing the power of Computer Vision to analyze customer movement and behaviors. These Heat Maps offer invaluable insights into high-traffic zones, customer preferences, and the strategic placement of products within retail spaces. Through the project, retailers were empowered to make data-driven decisions, optimizing store layouts and marketing strategies, ultimately elevating customer experiences and driving increased sales.

• Designed and implemented a collection and inference system;
• Orchestrated multiple threads to make human detections in real-time inside a low-resource device;
• Deployed System in Edge Device;
• Initiated and executed the creation of Heat Maps utilizing store security cameras;
• Employed Computer Vision techniques to analyze customer movement and behavior patterns, translating raw data into visual Heat Maps.