
Mary Newhauser developed and documented end-to-end Retrieval Augmented Generation (RAG) workflows for the weaviate/recipes repository, focusing on AI-enabled document processing and vector search. She built Jupyter notebooks demonstrating PDF parsing with Docling, embedding generation using ModernBERT, and integration with Weaviate for storage and retrieval. Her work included step-by-step guidance for installation, environment setup, and GPU configuration, as well as improvements to onboarding and documentation clarity. By addressing both feature delivery and bug fixes, Mary enabled rapid experimentation with machine learning and natural language processing, lowering barriers for teams to prototype, extend, and evaluate document retrieval solutions using Python.

December 2024 monthly summary for weaviate/recipes: Delivered an end-to-end ModernBERT embeddings with Weaviate integration recipe, including data loading, embedding generation, indexing into Weaviate, and sample queries. Also fixed Colab link path to ensure Colab notebooks are accessible for the ModernBERT embeddings recipe. Commits included: ab768bbfe383050c3fa21685cfef1e4ffdaae84d, 4356b7d6ef3f133e2a8136865e890b47ba4ba89b, and 274503e53a48a6ff08dc92237322f992d6ce6d97. Impact: lowers onboarding friction, enables rapid experimentation with vector search, and improves documentation clarity and discoverability. Technologies/skills demonstrated: Python, Colab workflows, ModernBERT embeddings, Weaviate vector search, data loading and indexing, documentation practices.
December 2024 monthly summary for weaviate/recipes: Delivered an end-to-end ModernBERT embeddings with Weaviate integration recipe, including data loading, embedding generation, indexing into Weaviate, and sample queries. Also fixed Colab link path to ensure Colab notebooks are accessible for the ModernBERT embeddings recipe. Commits included: ab768bbfe383050c3fa21685cfef1e4ffdaae84d, 4356b7d6ef3f133e2a8136865e890b47ba4ba89b, and 274503e53a48a6ff08dc92237322f992d6ce6d97. Impact: lowers onboarding friction, enables rapid experimentation with vector search, and improves documentation clarity and discoverability. Technologies/skills demonstrated: Python, Colab workflows, ModernBERT embeddings, Weaviate vector search, data loading and indexing, documentation practices.
November 2024 monthly summary for weaviate/recipes repository focusing on feature delivery and documentation improvements that enable rapid experimentation with RAG over PDFs. No major bugs fixed this period; emphasis was on end-to-end demonstration and environment readiness that unlocks business value through faster AI-enabled document workflows.
November 2024 monthly summary for weaviate/recipes repository focusing on feature delivery and documentation improvements that enable rapid experimentation with RAG over PDFs. No major bugs fixed this period; emphasis was on end-to-end demonstration and environment readiness that unlocks business value through faster AI-enabled document workflows.
Overview of all repositories you've contributed to across your timeline