Over a two-month period, contributed to the keras-team/keras repository by expanding OpenVINO backend support for advanced linear algebra operations. Developed and integrated the slogdet function using LU decomposition with partial pivoting, enabling sign and log-determinant calculations on Intel hardware. Refactored backend logic to unify implementations and improve code hygiene, which enhanced parity with NumPy and increased CI reliability. Further extended backend capabilities by implementing solve_triangular, ensuring correct type handling and seamless interoperability with Keras tensor operations. Leveraged Python, Keras, and OpenVINO to deliver robust, maintainable features that broadened deployment options and strengthened cross-backend consistency for machine learning workloads.