It’s no secret that cold startup times of traditional JVMs have given Java a bad reputation when it comes to running on Function-as-A-Service (FaaS) platforms like AWS Lambda. Lambda does cache instances of your functions, but whenever it has to instantiate a new one, the overhead of starting a new JVM instance can be significant. It doesn’t have to be this way. Oracle’s GraalVM can create native images with blazing fast startup times. Even though AWS Lambda doesn’t have built-in support for GraalVM, the two can be integrated using custom runtimes for AWS. Sure, GraalVM native images have some specific limitations, but depending on your use case, it may be worth it, especially for greenfield projects. In this session, we'll compare the performance of a simple Java function deployed normally, as a GraalVM native image, and with AWS SnapStart enabled. We’ll also take a quick look at one of GraalVM’s other killer features: support for polyglot development using Python, JavaScript, C/C++, and more.