Developed a SIMD-accelerated string search feature for the IBM/velox repository, focusing on enhancing text-processing performance in C++. The work introduced simdStrstr, a new method that leverages SIMD instructions to compare the first and last characters of the search needle in parallel batches before proceeding to full string comparison. This approach reduces unnecessary comparisons and improves the efficiency of search-heavy queries, particularly in scenarios where early character matches are common. The implementation replaced the traditional std::find in common search patterns, laying a foundation for broader performance optimization in Velox and demonstrating expertise in C++, SIMD, and string algorithms.