
Single data source
​​The model is independently trained with the new data 1, 2, 3 …


​​Integrate STADLE API into Model Training Process in the cloud or on-premise server


STADLE Orchestrator is to continuously train the model

Multiple data sources (cloud or edge training)
The models are independently trained with the new data 1, 2, 3 … on the edge and in the cloud


Integrate STADLE API into Model Training. Processes on the edge side and in the cloud


STADLE Orchestrator is to continuously train the models both on the edge and cloud

Edge training and inference together
The models are independently trained in different environments creating data silos …


Integrate STADLE API into both Model Training Processes and Deployment Processes in edge devices


STADLE Orchestrator aggregates and deploys models all in edge environments

Cloud training, edge deployment/inference
Manual deployment is required to update the edge device models


Integrate STADLE API into both Model Training Processes in the cloud and Deployment Processes in edge devices


STADLE Orchestrator aggregates models trained in the cloud and deploys models on edge

Subset edge training, subset edge inference
Model training processes are independent and manual deployment is required to update the edge device models


Integrate STADLE API into both Model Training Processes both in the cloud & edge and Deployment Processes in edge devices


STADLE Orchestrator aggregates models trained in the cloud and deploys models on some of the edge devices

Training through “virtual decentralization”, convert to decentralized case
By effectively Integrating STADLE API into centralized Model Training Processes in the cloud, efficient training of the model is realized


Data is dispersed into another training module to be orchestrated


Training modules can be extended as much as you want

