Worked on enhancing the stability and robustness of quadratic programming (QP) solutions in the cvxgrp/cvxpy-ipopt repository by focusing on the integration with the HiGHS solver. Addressed a critical issue in Hessian matrix formatting by switching from a square to a triangular representation, which resolved errors caused by asymmetry in floating-point calculations. This fix improved the numerical stability of QP solves and reduced solver errors. Contributed additional test coverage and documentation to ensure the reliability of these changes. Utilized Python for implementation, applying skills in algorithm design, numerical optimization, and testing to deliver a more robust solver integration workflow.