The employment and entrepreneurship career choice planning of college students is an important constituent module of the talent training system of colleges and universities in the new era. Aiming at the traditional ant colony algorithm with poor realm adaptability and a large number of inflection points, this paper proposes an ant colony algorithm based on Sigmoid statistical iteration. The Sigmoid activation function distribution strategy is adopted to reduce the blindness of the algorithm’s presearch, and the heuristic function is dynamically adjusted by the introduction of the adaptive factor to reduce the convergence time of the algorithm, and finally the pheromone update function is dynamically adjusted according to the number of iterations to construct the career choice path planning model and apply the model to the career choice planning path recommendation system. When the number of users is 1000, the average response time of the proposed system is only 322ms, the throughput is 394, and the pass rate is 100%, and the CPU occupancy and memory usage are lower than those of the traditional system (35.32% and 39.83%).