Having the ability to hear and talk are blessings that allow us to converse and communicate with each other with ease. However, due to some illnesses or disabilities, not all human beings are able to talk and hear. Some are born mute or deaf, and some lose these abilities because of illnesses or accidents. Luckily, to enable them to converse, sign language was created, where they rely on their kinetic and visual abilities to interpret what they want to say or to understand what is said to them. Yet, in this case, instead of spoken words, signs are used. Although sign language is the bridge connecting the deaf and mute communities together and with normal people, it is still only learned by those who need to use it frequently. Limiting the chances of people from these communities to converse with ease with a wider range of normal people. In this paper, we explain the development process of Talk2Me which is a real-time sign language translator mobile app using computer vision and deep learning techniques. Talk2Me was developed to extend the bridge connecting the deaf and mute communities to a wider range of normal people without the need to learn sign language. In its initial version, we used transfer learning on the YOLOV5S model to recognize the Algerian sign alphabets and numbers. The Algerian sign alphabets and digits dataset images were self-collected, labeled, and preprocessed. Talk2me now provides real-time translation of Algerian signs in textual and oral formats.