Gesture Recognition plays a crucial role in facilitating and enhancing communication accessibility for individuals with hearing and speech impairments. However, translating complex sign language into spoken or written language remains a significant challenge. In an effort to address this, this research utilizes the MediaPipe framework and the Random Forest Classifier algorithm to classify sign language gestures and expressions in BISINDO (Indonesian Sign Language). Considering the difficulty and complexity of sign language gestures, 10 expressions/words in BISINDO were selected, resulting in a dataset of 25,000 data points used in this study. The approach involves detecting sign language through pose, hand, and facial gesture or movement recognition. Evaluation results show that the Random Forest algorithm achieves a remarkably high level of precision, recall, F1-score, and accuracy (99.88%). Additionally, the developed system demonstrates good performance with prediction probabilities ranging from 0.50 to 0.70 for correct predictions, although challenges persist in distinguishing similar sign gestures, resulting in some predictions requiring more time to yield accurate results. The findings of this research contribute significantly to improving sign language recognition and promoting inclusivity for individuals with hearing and speech impairments. Moreover, it opens up new opportunities for further advancements in sign language detection technology.