Ønisa Jr

Developed a system to analyze GPS trajectory data for estimating foot traffic patterns and impression exposure, visualizing routes with Dash and Plotly. Engineered machine learning pipelines for large-scale spatial-temporal datasets to support predictive analytics for urban mobility and marketing attribution.

Optimized computer vision algorithms for crowd density counting, reducing Mean Absolute Error by 22% while maintaining real-time performance. Implemented transfer learning and edge computing solutions to improve accuracy and reduce cloud costs.

Developed transformer-based algorithms for crowd counting and unsupervised clustering systems for power grid disturbances. Led computer vision projects with perfect client ratings and conducted technical interviews for data science positions.

Designed and implemented ML systems for wildlife detection with 92% accuracy, deployed across web platforms and field monitoring stations. Improved pipeline efficiency by 40% and reduced downtime by 25% through optimization and monitoring strategies.

Conducted customer behavior analysis using Python, NoSQL, and statistical methods, improving prediction models by 18%. Developed LSTM neural networks for time-series forecasting with 87% accuracy and created data preprocessing pipelines that reduced training time by 30%.

Pre-Bachelor in Mechatronics and Robotics Automation with coursework in machine learning, computer vision, embedded systems, robotics control, and sensor networks. Capstone project on automated precision agriculture using computer vision and IoT.