Update readme

README.md

@@ -10,13 +10,23 @@
 [![Go Report Card](https://goreportcard.com/badge/github.com/dpogorzelski/speedrun)](https://goreportcard.com/report/github.com/dpogorzelski/speedrun)
 [![Go](https://github.com/dpogorzelski/speedrun/actions/workflows/go.yml/badge.svg)](https://github.com/dpogorzelski/speedrun/actions/workflows/go.yml)
 
-Speedrun executes commands, at scale.
+Speedrun is a command execution framework that works at scale.
+The point of Speedrun is so that you can run any command across any number of servers, projects and cloud vendors with ease and fast (currently GCP only but AWS and Azure will be supported as well), example:
+`speedrun run systemctl stop nginx` to stop nginx across 3k machines.
+
+No hassels with setting up and maintaining a server with agents as speedrun has none. No runtime to install/maintain as it's a single self-contained binary. Speedrun leverages SSH as transport with tight cloud vendor integration to take the burden of mundane things like key generation and propagation/revocation away from the user.
+
+Server selection is made as intuitive as possible, currently based on [gcloud topic filters](https://cloud.google.com/sdk/gcloud/reference/topic/filters) but in the future will be replaced by a generic selection mechanism to provide a seamless experience across different cloud vendors.
+
 
 Features:
 
 * native cloud integration (currently Google Cloud only, AWS and Azure coming up!)
 * stateless and agentless
 * no complex configuration required
+* single self-contained binary
+* can run against any number of servers and projects seamlessly
+* a plugin system is in the plan to allow anyone to integrate execution modules that will wrap complex functionality instead of running raw shell commands
 
 
 ## Installation