During August 2025, Mhowto focused on integrating Gemini embedding capabilities into both the cloudwego/eino-ext and coze-dev/coze-studio repositories. Working primarily in Go, Mhowto developed a new backend component for embedding generation, implementing configuration management and embedding logic to support Google’s Gemini models. The integration included environment variable handling and basic test coverage, ensuring reliable initialization and consistent deployment across cloud services. By standardizing embedding configuration and logic, Mhowto enabled smoother cross-team adoption and laid the foundation for downstream machine learning features such as improved search and similarity. The work demonstrated depth in API integration and embedding model implementation.