diff --git a/docs/blog/2024-02-01-grime-guardian.md b/docs/blog/2024-02-01-grime-guardian.md
index a92d7549..c880d3e4 100644
--- a/docs/blog/2024-02-01-grime-guardian.md
+++ b/docs/blog/2024-02-01-grime-guardian.md
@@ -41,7 +41,7 @@ The Grime Guardian demonstrates how to build an advanced Groundlight application
 1. **Raspberry Pi Deployment** - The Grime Guardian leverages our custom [Raspberry Pi Image](https://github.com/groundlight/groundlight-pi-gen), which makes it easy to deploy Groundlight applications on Raspberry Pi. 
 2. **Multiple Cameras** - The Grime Guardian actively uses more than one camera to solve a problem (it has one camera pointed at the sink and one pointed at the general kitchen scene). 
 3. **Multiple Detectors** - The Grime Guardian combines multiple Groundlight detectors to solve a problem.
-4. **Microservices-like process via Multithreading** - The Grime Guardian’s architecture is broken down into a handful of microservice-like processes - each running in a different thread on the same machine. This improves the app’s robustness and allows for more flexibility and scalability. 
+4. **Microservice-like architecture via Multithreading** - The Grime Guardian’s architecture is broken down into a handful of microservice-like processes - each running in a different thread on the same machine. This improves the app’s robustness and allows for more flexibility and scalability. 
 5. **Complex State** - As described in the previous section, the state of the world this app is tracking is somewhat complex. In addition to knowing the state of the sink and kitchen, the app tracks how recently the state was updated and how recently it has sent a notification to the Groundlight team. 
 6. **Discord Bot Integration/Notifications** - The Grime Guardian uses the Discord Bot API to send notifications to a Discord server. Discord can be an extremely powerful and flexible tool for building applications (e.g. Midjourney). 
 7. **Robustness** - In practice, the Grime Guardian has been extremely robust, with only one or two incorrect (false positive) notifications over many weeks of deployment and hundreds of thousands of Groundlight queries.