In May 2025, Diadorer delivered end-to-end Nebius feature extraction support for the Hugging Face Hub ecosystem, working across the huggingface/huggingface_hub and hub-docs repositories. They implemented a new feature extraction task for the Nebius inference provider, integrating it into the Hub library using Python and Markdown for both code and documentation. This work aligned technical documentation with the updated codebase, clarifying Nebius capabilities and improving user onboarding. Diadorer’s contributions focused on API integration and full stack development, enabling users to leverage Nebius for feature extraction workflows and expanding available inference provider options without addressing major bug fixes.