Skin related issues are increasing due to changing lifestyles, environmental conditions, and lack of proper skincare awareness. Many individuals struggle to accurately identify their skin type and select suitable skincare products. Selection and use of incorrect or unsuitable products can lead to various skin problems. With the rapid advancement of AI, it has become feasible to develop smart skin recommendation systems by using AI techniques. This emphasizes the need for an intelligent and automated solution that can deliver personalized skincare guidance based on skin type and other user inputs. In this work, we propose Smart-Derma, an intelligent skincare recommendation system that leverages deep learning techniques to analyze skin conditions and provide personalized suggestions. In the proposed system, firstly, facial skin images are acquired from the user. The images are preprocessed to remove noise and extract important features. Deep learning models are then applied for analysis. CNN is used for skin type classification and YOLO is used to detect specific skin issues. The detected features and other user inputs are further processed using a hybrid recommendation technique. This recommendation system combines both rule-based and content-based techniques to generate personalized skincare suggestions. The proposed system achieves better performance than existing methods, especially in terms of higher accuracy and effective recommendations.