Early and precise diagnosis of dermatological problems can be challenging, especially in regions with limited access to specialists. This paper presents an AI-based diagnostic tool that analyzes skin condition images using Random Forest techniques and deep learning. Trained on a diverse dataset, the system can accurately detect and classify conditions such as Psoriasis, Seborrheic Dermatitis, Lichen Planus, Pityriasis Rosea, Chronic Dermatitis, and Pityriasis Rubra Pilaris. It provides fast and reliable diagnostic support, minimizes human error, and aids medical professionals in early intervention. Advanced image preprocessing ensures consistent performance across varying image qualities, while strict adherence to data privacy standards (HIPAA, GDPR) protects patient information. Validation results show high accuracy (94.8%), precision (92.5%), and recall (93.6%). Compared to other AI models, the system demonstrates improved diagnostic performance and reduced processing time. This study outlines the architecture, methodology, validation results, and limitations of the tool, highlighting its potential to enhance accessibility and effectiveness in dermatological diagnostics.