Predicting CAD Metadata with the HOOPS AI Context Layer

These notebooks combine HOOPS Embeddings similarity search with metadata from neighboring parts to predict missing information about a query part.

Follow Install HOOPS AI for environment and license setup, and build the search index described in CAD Indexing and Retrieval Using HOOPS Embeddings before starting.

Work through the notebooks in this order:

  1. demo_HOOPS_Embeddings_Context_Layer_quickstart.ipynb introduces the end-to-end workflow: load the model and index, connect a metadata store, retrieve similar parts, and predict missing fields.

  2. demo_HOOPS_Embeddings_+_Context_Layer.ipynb customizes that workflow with aggregation rules, metadata normalization, known query context, review policies, cost estimates, and checks for mislabeled parts.

The code and resources for this tutorial can be found in the HOOPS-AI-Tutorials GitHub repository.

Info

From the repository root, run jupyter lab embeddings_pipeline/demo_HOOPS_Embeddings_Context_Layer_quickstart.ipynb to start with the quickstart before exploring customization.