Developed a configurable AI model interaction feature for the firebase/genkit repository, focusing on the Ollama plugin to enable runtime control over model parameters such as temperature and reasoning capabilities. This work leveraged Go for plugin development and API integration, allowing users to adjust AI model behavior dynamically to better suit diverse use cases. By introducing runtime configuration options, the solution reduced iteration time and laid the foundation for more flexible and cost-effective AI workflows. The approach emphasized business value by supporting broader deployment scenarios and optimizing performance, while also involving collaborative contributions to ensure robust and maintainable plugin architecture.