SingleStore has introduced an APPROX_PERCENTILE function to estimate percentiles for a distribution of values, providing a faster and near-exact alternative to traditional percentile functions like PERCENTILE_CONT. This new function leverages the performance benefits of SingleStore's database and uses the T-digest algorithm to achieve accurate results with minimal data sorting. By using APPROX_PERCENTILE, users can speed up their percentile calculations without sacrificing accuracy, making it an ideal solution for analyzing large datasets where exact calculations would be too slow or impractical. The function offers adjustable accuracy parameters and is designed to work well on massive datasets, providing a reliable and efficient way to calculate percentiles in real-world applications.