Kien Nguyen developed a foundational differential privacy feature for the IBM/velox repository, focusing on secure analytics in Presto SQL. He implemented the NOISY_COUNT_IF_GAUSSIAN aggregation, which counts TRUE values in boolean columns while adding Gaussian noise to preserve privacy. Using C++ and SQL, Kien ensured the aggregation’s results remain non-negative bigints through careful post-processing, maintaining query correctness under differential privacy constraints. This work established a core capability for privacy-preserving analytics in Velox, aligning with product goals for secure customer deployments. The depth of the implementation reflects strong skills in aggregate functions, data engineering, and privacy-aware database development.