Cold starts in serverless functions can be caused by various factors such as new function containers being initialized, unneeded dependencies, and tightly scoped module imports. Datadog Serverless Monitoring detects cold starts, visualizes their impact on services through distributed traces, and allows for alerts based on the rate they occur. Cold Start Tracing via Datadog Serverless APM provides an under-the-hood view of dependencies loaded during a cold start, enabling developers to identify root causes and gain actionable insights to optimize performance and reduce costs. This feature helps developers visualize the steps executed during a function's cold start, such as downloading libraries and modules, and enables them to go beyond simple detection by identifying which parts of their code are contributing to cold starts. By using Cold Start Tracing, developers can fine-tune their Lambda function configurations, test changes, and improve the performance of their serverless functions with support for Node.js and Python runtimes coming soon.