This paper presents a genetic algorithm (GA) method to build the load-settlement curve of a driven pile in glacial deposits. The load-settlement mechanism of an axially loaded piles is a complex soil-structure interaction problem and influenced by many factors, including the pile dimensions, installation method, and ground conditions. Due to these uncertainties, pile load tests have been used for decades to verify the design assumptions in Ontario, Canada. A database of more than 168 pile load tests was collected by the Ministry of Transportation of Ontario (MTO). In this study, a simple hyperbolic curve is adopted to simulate the load-settlement curves of 36 pile compression tests in various parts of Ontario. GA, one of artificial intelligence techniques, is applied to correlate the soil and pile parameters to the two hyperbolic parameters. Based on pile and ground conditions, the proposed formulas can be used to build a load-settlement curve for the serviceability design of a pile driven in glacial deposits.