Shreyas delivered quantization support for the Eagle draft model in the IBM/vllm repository, focusing on efficient model execution and deployment. Using Python and leveraging skills in machine learning and model optimization, Shreyas implemented an end-to-end quantization flow integrated directly into the model architecture. The work included developing unit tests to validate quantization configurations and documenting configuration paths to enable flexible tuning across Eagle model variants. This engineering effort improved inference performance and reduced memory usage for draft Eagle models, addressing scalability for production environments. The depth of the work is reflected in the comprehensive integration and validation of quantization features.