Serverless abstracts infrastructure management, enabling event-driven, stateless services that you only pay for when used, and its history can be understood in four distinct phases.
Google App Engine in 2008 was a precursor to serverless, offering a fully managed platform for web apps but with significant limitations in language support and ecosystem.
The introduction of AWS Lambda and Google Cloud Functions popularized FaaS but came with challenges like limited triggers, short execution times, and the infamous cold start problem.
Serverless container platforms like Google Cloud Run brought language flexibility, solved the cold start problem with higher concurrency, and enabled better local development workflows.
Services can be connected indirectly via event-driven choreography for loose coupling or managed directly by an orchestrator like Google Workflows for clear, traceable logic.
Advanced architectures often combine event-driven choreography with orchestration and can even integrate serverless workflows with long-running tasks on traditional virtual machines.
With its maturity in triggers, resources, and orchestration, serverless has become the default, cost-effective, and maintainable approach for running code in the cloud.
28:55Ingo Weichsel
30:03Marco Plaul & Martin Sakowski
42:23Adam Bien
46:24David Leitner
1:01:02Oliver Arafat
23:19Nočnica Fee
41:13Cheng Zhang
27:11Kevin Dubois & Daniel Oh
.webp?w=240&auto=compress,format)
Jobs that call for the skills explored in this talk.
doinstruct Software GmbH
Berlin, Germany
Peter Park System GmbH
München, Germany
CONTIAMO GMBH
Berlin, Germany
SYSKRON GmbH
Regensburg, Germany
AUTO1 Group SE
Berlin, Germany
