Secure multi-party computation (MPC) enables multiple parties to jointly evaluate a function on their private inputs without revealing any information beyond the output. It is a foundational tool for privacy-preserving applications, including federated analytics, private machine learning, distributed credential issuance, and differentially private data release. Despite decades of progress, a persistent gap remains between the theoretical efficiency of MPC protocols and the performance and usability demands of real-world deployments.

My work is motivated by a core goal: to build secure computation frameworks that are \emph{provably efficient} while remaining \emph{practical for real-world developers}. I work across three areas: (1) generic MPC protocol design, (2) compiler frameworks for MPC, and (3) application-driven cryptographic systems, including differential privacy and thresholdizing standardized post-quantum signature schemes. I propose to build an efficient threshold version of the NIST-standardized FALCON signature scheme along two complementary directions: FALCON-specific algorithmic improvements, including a fixed-point analysis of the fast Fourier orthogonalization sampler to reduce the asymptotic signing complexity, and improved MPC primitives, including more efficient correlation generators applicable to both FALCON and ML-DSA.