Worked on stabilizing the deployment process for the roboflow/inference repository by addressing startup reliability issues in slim Docker images. Focused on dependency management and DevOps practices, the developer identified that missing system utilities were causing failures in constrained environments. By adding the psutil package to the Dockerfile requirements, they ensured that necessary system resources were available at startup, directly reducing deployment errors and improving production uptime. This work demonstrated a disciplined approach to containerization and Python packaging, leveraging Dockerfile expertise to enhance service reliability and maintainability in resource-limited scenarios, with a clear focus on operational stability and robust deployment workflows.