This architecture allows for a number of advantages, such as:
Efficiency
The cloud can train and update the localized intelligence on the edge, which can then be used to perform
tasks locally without having to send data back to the cloud. This can save time and bandwidth.
Flexibility
The cloud can be updated with new data and models as needed, which can improve the performance of
the localized intelligence on the edge. This makes it possible to adapt the AxiPHYL Edge Assistant to
changing conditions and requirements.
Scalability
The cloud can be scaled up or down as needed, which can help to ensure that the AxiPHYL Edge Assistant
can handle any amount of traffic. This makes it possible to use the AxiPHYL Edge Assistant for a wide range
of applications, from small personal devices to large enterprise systems.
Overall, a distributed AI architecture with a large model on the cloud and localized intelligence on the
edge is a powerful and flexible approach to AxiPHYL Edge Assistant s that can be used to improve the user
experience, increase security, and reduce costs.
AxiPHYL is a large healthcare system in India that is using distributed AI to improve patient care. The
system’s architecture consists of three main components:
The cloud: The cloud is responsible for training and inference on large datasets of data. This allows the
AxiPHYL Edge Assistant s to health workers to learn a wide range of skills, such as voice recognition, natural
language understanding, and question answering. The cloud also stores and manages data, such as user
preferences and device information. This data is used to personalize the AxiPHYL Edge Assistant
experience and improve the accuracy of its responses.
The edge (mobile, tab, kiosk or device): The edge is responsible for real-time tasks, such as video, image,
voice recognition, responses and extraction of data and get medical response. This requires low latency,
so the edge is mobile and typically located with the user. The edge also has increased privacy and security,
as it does store local patient data for easy access but it’s secured. It works as second brain for Health
Workers.
The user (Health Worker): The user is the person who interacts with the AxiPHYL Edge Assistant. The user
can use video, image, voice, speech or text input to interact with the AxiPHYL Edge Assistant. The AxiPHYL
Edge Assistant then generates a response to users or patients questions.
Here is an example of how AxiPHYL architecture is used to provide real-time assistance to doctors and
nurses in the primary care centers. When a health workers use the inbuilt AxiPHYL assistant. The AxiPHYL
Edge assistant then uses its knowledge of the patient, questions, answers based on language and dialect
and the procedure, and the medical equipment to provide assistance. For example, the AxiPHYL assistant
can help the doctor or nurse to identify and correct errors, and it can also provide information about the
patient, the procedure, and the medical equipment.