Doc In A Pock

Mobile Application

About the project

Usage of our application allow users can take medical diagnostics and receive links to the verified articles with possible diagnoses. The user reports initial symptoms to a special neural network in the chat mode, application generates additional questions and establishes a diagnosis. The neural network has been trained on more than 30 million medical cases and articles; on tests, Neural network’s diagnosis accuracy determination has exceeded 80%. User can show the diagnostic results to his general practitioner or read the proposed medical articles and contact a focused expert.

UI/UX research and designMobile App: Flutter/Dart


Client’s task was to make a simple and convenient product for its users. Our company has developed a concept and a customer journey within the product turning a step-by-step assistant with questions into a convenient chat. Therefore interaction with the application has become easier and more like a live chat with an online consultant. The product was then packaged into a mobile app for iOS and Android and released to the Apple and Play stores.


User’s first application experience will be about: application description, usage guide and app’s key advantages. Onboarding describes competent AI performance and illustrated learning in a variety of medical articles. Onboarding immediately warns user, do not fully rely on the application, but to only take the results of the diagnosis and add it during the doctor appointment.



The application contains question database which helps AI assistant to suggest a diagnosis. In the initial survey users report their gender, age, and allergies. The application generates clarifying questions based on the primary data provided by the user and based on the answers suggests a possible diagnosis.



The main function of the assistant is to make a diagnosis based on the symptoms and reports received from the user. The assistant itself offers additional symptoms in the form of questions which user can confirm or deny. After collecting enough information, the assistant offers possible diagnoses and links to their detailed description. Assistant always offers the user to read medical articles and recommends that you contact a doctor with the information received after possible diagnosis determination. In addition, the assistant contains an algorithm that analyzes the user's symptoms connected with a direct life risk. If symptom’s series coincide the assistant may advise you to immediately contact 911 and tell what happened to you.



User has full access to the history of chats with the assistant where you can find and select each appointment separately and look the whole conversation chain up. The case record as well as individual conversations can be downloaded as PDF files for your doctor consideration.