diff --git a/README.md b/README.md
index e608db5..c04b480 100644
--- a/README.md
+++ b/README.md
@@ -6,9 +6,12 @@ A basic example of microservice architecture which demonstrates communication be
 * Uses RabbitMQ to communicate between services
 * Uses WebSocket to talk to the front end
 * Stores data in PostgreSQL
+* Stores cache in Redis
 * Uses React for front end development
-* Uses Redis for cache
 * Builds and runs with Docker
+* Deployed on AWS with CloudFormation templates
+    * ECS using EC2
+    * AWS Fargate
 
 ![](demo.gif)
 
@@ -24,14 +27,6 @@ cd microservices
 docker-compose up
 ```
 
-### Back end
-
-To access the back end service, attach to its docker container from a separate terminal window. Messages from the front end will show up here. Also, standart input will be sent to the front end for two way communication.
-
-```bash
-docker attach microservices_backend
-```
-
 ### Database
 
 To inspect the database, launch a new container that will connect to our Postgres database. Then enter the password `demopsw` (see the `.env` file).
@@ -49,21 +44,80 @@ Select everything from the messages table:
 select * from messages;
 ```
 
+### Redis
+
+To inspect Redis, connect to its container via redis-cli.
+
+```bash
+docker run -it --rm \
+    --network microservices_network \
+    redis:6-alpine \
+    redis-cli -h redis
+```
+
+Get all cached messages or show the number of messages.
+
+```bash
+lrange messages 0 -1
+get total
+```
+
 ### RabbitMQ
 
 Access the RabbitMQ management interface by visiting `localhost:15672` with `guest` as both username and password.
 
-### Redis
+### Back end
+
+To access the back end service, attach to its docker container from a separate terminal window. Messages from the front end will show up here. Also, standart input will be sent to the front end for two way communication.
 
 ```bash
-docker run -it --rm --network microservices_network redis:6-alpine redis-cli -h redis
+docker attach microservices_backend
 ```
 
+## Deployment on Amazon ECS/AWS Fargate
+
+### ECR
+
+To deploy on ECS, the first step is to publish the service images to Amazon ECR. In Makefile, replace the `registry` variable with your own registry. [Login](https://docs.aws.amazon.com/AmazonECR/latest/userguide/registry_auth.html#get-login-password) to ECR. Then run the following command:
+
 ```bash
-lrange messages 0 -1
-get count
+make ecr
 ```
 
+### CloudFormation stacks
+
+There are several CloudFormation stacks to create. `cd deployments` and create the network stack:
+
+```bash
+aws cloudformation create-stack --stack-name MicroservicesNetwork --template-body file://network.yml
+```
+
+Create the following stacks in any order (for an EC2 cluster). To create a Fargate cluster, change the `cluster-ec2.yml` to `cluster-fargate.yml`.
+
+```bash
+aws cloudformation create-stack --stack-name MicroservicesResources --template-body file://resources.yml
+aws cloudformation create-stack --stack-name MicroservicesAlb --template-body file://alb.yml
+aws cloudformation create-stack --stack-name MicroservicesClusterEC2 --template-body file://cluster-ec2.yml --capabilities CAPABILITY_NAMED_IAM
+```
+
+Once done, create the services (for an EC2 cluster). For Fargate, change the `services-ec2` directory to `services-fargate`.
+
+```bash
+aws cloudformation create-stack --stack-name MicroservicesServiceServer --template-body file://services-ec2/server.yml
+aws cloudformation create-stack --stack-name MicroservicesServiceCache --template-body file://services-ec2/cache.yml
+aws cloudformation create-stack --stack-name MicroservicesServiceDatabase --template-body file://services-ec2/database.yml
+```
+
+To visit the website, head to the MicroservicesAlb stack and open the `ExternalUrl` from the Outputs tab.
+
+### Deleting stacks
+
+When deleting the `cluster-ec2.yml` stack in CloudFormation, delete the auto scaling group manually from the AWS EC2 console.
+
+### References
+
+Deployment is based on these templates: https://github.com/nathanpeck/ecs-cloudformation
+
 ## Development environment
 
 For development, run the RabbitMQ and Postgres containers with Docker Compose.