UiPath AI Center provides flexibility for integrating third-party ML models into its framework. Here's how:
Third-Party Model Integration: Users can create a custom package for AI Center that incorporates a third-party model. This involves exporting the model in a compatible format (e.g., ONNX or TensorFlow SavedModel) and wrapping it into a package deployable in AI Center.
Scenarios for Usage: This is especially useful when businesses already have proprietary models developed externally or sourced from other vendors. These can be fine-tuned and deployed alongside UiPath's RPA workflows.
Infrastructure Consideration: While deploying on-premises provides better control over model usage and performance, AI Center supports GPU acceleration and third-party model integration across different deployment modes, ensuring scalability and processing efficiency.
This approach allows businesses to maintain the versatility of their ML pipelines and integrate advanced analytics with minimal disruption to their existing automation setup.