During January 2026, Bhaskar Bonik developed a Jupyter notebook for the aws-samples/amazon-bedrock-samples repository, enabling seamless import and serverless inference of fine-tuned Qwen3 models into Amazon Bedrock using the Custom Model Import workflow. He focused on building an end-to-end solution in Python that guides users through environment setup, artifact download from HuggingFace, and inference steps, all within a reproducible notebook format. Leveraging AWS and machine learning expertise, Bhaskar’s work addressed the need for streamlined model experimentation by establishing a reusable pattern for future imports. The project emphasized documentation and feature completeness, with no major bug fixes required.