I am a Data Scientist and fairly new to using Dapr. I find it to be an incredibly powerful framework, especially for someone like me who is more focused on solving problems than on perfecting code. The Dapr sidecar model significantly reduces boilerplate, allowing me to quickly develop AI microservices and hand them off for testing.

Currently, I deploy using a docker-compose.yml file that includes both Dapr-based and non-Dapr code. The deployment to production is managed through Azure DevOps pipelines, and we use Grafana for monitoring. This setup works well for us, but I’m curious about the prevalent use of Kubernetes for Dapr deployments. It seems unnecessarily complex and introduces additional dependencies for developers, yet it appears that most Dapr developers prefer Kubernetes for production.

I would greatly appreciate any insights or feedback on this topic. Why is Kubernetes so commonly used for Dapr deployments? Are there specific benefits or use cases where Kubernetes is essential, or is it just a matter of preference?

I have already built and developed my dapr solution using docker compose and it is working fine. I am curious what incremental value you get from k8