This work presents a hybrid algorithm for designing tourist routes at destinations that combines two metaheuristics: Greedy Randomised Search Procedure (GRASP) and Variable Neighbourhood Search (VNS). The search for high-quality solutions maximises the sum of the scores (preferences) of the visited points of interest, considering the service characteristics and time limitations. This approach is particularly suitable for tourists planning optimal routes to visit attractions. Users can easily add or remove visits, adjust preference scores, and consider other restrictions. The goal is not to compete with state-of-the-art algorithms for standard tourist trip design problems (TTDP) using classical benchmarks. However, preliminary experiments with these benchmarks show acceptable results. The research aims to develop a component for tourist route recommendation systems that efficiently search for high-quality routes.