Ansys SimAI Workflow

The following illustration depicts the Ansys SimAI platform user workflow. The main actions described in the picture are explained with more details in the paragraphs below with links to the relevant parts of the documentation.

• Data Preparation: To be as agnostic as possible, the Ansys SimAI platform only ingests neutral formats as input. Simulation data coming from your solvers must be exported or converted into that format. To make sure your data are well-formed and compliant with our best practices, the platform guides you through the expected format before uploading it into the platform. For more information, see Training Data Preparation.
• Data Upload, and Model Configuration and Building: Once your data are uploaded, you can move onto the AI model building process. Consult Model Building to cover all the steps related to the creation of your AI model. From training data selection to model settings definition and model building.
• Prediction, and Postprocessing: Once your AI model is built, you can use it to perform predictions on your design geometries. Some postprocessing actions are also available to you to either get a quick insight and indicators on your model's performance or to download relevant outputs for further postprocessing outside of the Ansys SimAI platform. For more information, see Prediction.
• Analysis, and Validation: Use our Model Evaluation Report and the other postprocessing available to analyze your model, identify potential failing designs and how to alter your model settings or training data to improve performance. When you have a model that performs as you expect it, validate your best performing designs with full fidelity simulation. For more information, see Analysis.