Developed and delivered a configurable embedding batch size feature for the lfnovo/open-notebook repository, enabling support for diverse embedding providers, including CPU-only environments and OpenAI-compatible endpoints. The implementation leveraged Python for backend development and focused on robust environment configuration, allowing batch size to be set via an environment variable. This approach addressed provider-specific constraints and facilitated easier experimentation with cost and performance optimization. The work included careful validation and adjustments based on code review feedback, enhancing configuration reliability. By establishing this flexible infrastructure, the project laid the groundwork for broader cross-provider embedder support and improved deployment flexibility across environments.