Contributed to the PyMC and PyTensor repositories by expanding modeling capabilities and improving reliability in scientific computing workflows. Developed Dirichlet distribution and SimplexTransform support within pymc.dims, enabling more robust handling of simplex-constrained multivariate distributions using Python and probabilistic programming techniques. Enhanced the xtensor module in PyTensor by implementing a logsumexp function, validated for numerical accuracy against SciPy, to improve tensor computation stability. Addressed documentation accuracy by refining deprecation warnings in batched_dot, reducing migration friction for users. Streamlined test suites for censored distributions, simplifying maintenance and improving test reliability. Work emphasized Bayesian modeling, numerical analysis, and comprehensive software testing.