Numerical Error Analysis
Floating-point error, backward and forward error, sensitivity, conditioning, and rounding-error propagation.
Mathematical reliability and verified computation
Senior Scientific Reviewer | Computational Mathematician
Strengthening calculator reliability through numerical error analysis, high-precision reference checks, convergence testing, and transparent algorithm assumptions.
Dr. Sophia Laurent is a computational mathematician and senior scientific reviewer specializing in numerical error analysis, high-precision algorithms, interval arithmetic, optimization, root-finding methods, verified computing, and trustworthy mathematical software.
Her work connects rigorous numerical analysis with practical software verification, with attention to formula correctness, floating-point behavior, convergence testing, edge-case performance, and reproducibility of computational outputs.
At the Calzivo Research Review Board, she focuses on independent mathematical validation and numerical robustness assessment for online calculators and computational tools.
Dr. Laurent contributes senior scientific review for Calzivo through independent formula validation, numerical algorithm verification, tolerance assessment, convergence review, high-precision reference testing, and documentation review for calculator and computational-tool workflows.
Mathematical and software reliability strengths from the approved CV.
Floating-point error, backward and forward error, sensitivity, conditioning, and rounding-error propagation.
Interval methods, high-precision arithmetic, reference cases, and verified computation workflows.
Formula reconstruction, symbolic/numeric consistency checks, regression tests, and reproducibility dossiers.
Nonlinear solvers, numerical optimization, adaptive quadrature, convergence diagnostics, and tolerance review.
Input domains, assumptions, boundary behavior, adversarial inputs, and user-facing caveats.
NumPy, SciPy, SymPy, mpmath, Arb, MPFR workflows, interval arithmetic libraries, and MATLAB Symbolic Math Toolbox.
IRR/APR-style calculations, derivative-pricing formulas, optimization routines, and convergence checks.
Versioned formulas, source audits, data audits, reviewer sign-off, and clear limitations.
Leads independent mathematical verification of online calculators and computational tools, assessing correctness, numerical robustness, convergence behavior, tolerances, edge cases, and reproducibility.
Led research in numerical reliability and algorithm certification, with emphasis on trustworthy scientific software and validated computation.
Worked on certified numerical computing, verifiable algorithms, automatic differentiation, and interval arithmetic for robust computational tools.
Conducted doctoral research on self-verified numerical methods for reliable scientific computing.
Education and credentials listed from the approved CV.
EPFL, 2009 - 2013. Dissertation: Self-Verified Numerical Methods for Reliable Scientific Computing.
ETH Zurich, 2007 - 2009. Thesis: Error Propagation and Stability Analysis in Iterative Algorithms.
University of Geneva, 2004 - 2007. Graduated Magna Cum Laude.
Dr. Laurent contributes mathematical review to Calzivo where algorithm choice, numerical reliability, error behavior, reproducibility, and user-facing assumptions need expert validation. This profile does not claim that every Calzivo calculator has been individually reviewed by her.
Profile information is provided for professional transparency. Calzivo calculators are practical tools and do not replace professional advice.