[Webinar] Build Your GenAI Stack with Confluent and AWS | Register Now
Leverage GoodLabs and Confluent's data streaming platform for a virtual training platform, developed as a resource to allow medical students to practice their clinical examination skills by interacting with an AI-powered virtual patient.
In recent years, there has been a significant decrease in clinical skills within the field of medicine. The reliance on tests and imaging has grown, leading to more time spent on administrative tasks behind a computer. As a result, medical professionals are spending less time in direct interaction with patients, and when they do, their clinical skills are often found to be lacking.
Medical education programs assess clinical skills using the Objective Structured Clinical Examination (OSCE), yet there is a scarcity of resources focused on teaching these crucial clinical skills to medical learners. Our platform addresses this gap by offering medical learners the opportunity to refine their clinical skills through interactions with a virtual AI patient. It provides access to an extensive array of AI-patients, each with a medical history and details that have been thoroughly tested and validated for accuracy. This innovative solution is beneficial not only to students entering the medical field but also to practicing medical professionals seeking to improve their existing skill sets.
This platform is ideally suited for institutions involved in the training and employment of medical professionals, such as universities and healthcare providers, including hospitals, healthcare clinics, and urgent care centers. Universities can leverage this solution to give their students easy access to study materials, helping them prepare for exams and hone their clinical skills. Healthcare institutions can utilize it as a training tool for the orientation of new medical staff and as a regular training mechanism for their employees to ensure their clinical skills remain sharp.
Ultimately, this platform is designed with the future well-being of patients in mind. By enhancing the clinical skills of medical professionals, we can significantly improve the quality of care that patients receive.
The virtual training platform was developed as a resource to allow medical students and professionals to practice their clinical examination skills by interacting with an AI-powered virtual patient.
The AI platform provides medical students with extensive medical scenarios to practice their clinical skills based on the Medical Council of Canada.
The platform provides live interaction with highly specialized AI patients to respond and present in a given medical scenario, including a synchronous avatar linked to a speech generator.
The platform automatically assesses and provides feedback to the medical learner. Each scenario is recorded and tracked for the user’s review.
The platform offers AI patients exhibiting a range of clinical conditions, differing in severity, age, and associated morbidities.
Our physicians designed and evaluated the platform thoroughly to ensure authentic patient interactions and assessments.
Each AI patient can engage in realistic, real-time, two-way conversations with the medical learners, enhancing the training experience.
This use case leverages the following building blocks in Confluent Cloud:
Medical students, interns, or other medical professionals who want to enhance their clinical skillset.
Web conference app platform for medical professionals to practice their clinical skillset.
Websocket integrated with Confluent to bring physician audio in WebM into a Confluent topic.
Flink running speech-to-text AI to convert streaming caller audio into streaming speech and send it to a Confluent topic haystack.vishing.caller.speech.
Flink running a LLM to receive data from haystack.vishing.caller.speech along with haystack.vishing.caller.analysis to generate a dialogue responding back to the caller as a conversation agent and send the response to a Confluent topic haystack.vishing.agent.speech.
Integrates with Heygen Streaming API to feed the medically trained LLM response to generate the streaming AI patient avatar.
Contact GoodLabs to learn more about this use case and get started.