Deploy to AWS

Deploy AgentOS to AWS with ECS Fargate, RDS PostgreSQL, and a public endpoint.

This template deploys AgentOS and PostgreSQL to AWS. It includes example agents ready to use after deployment, managed infrastructure, and a public endpoint.

Prerequisites

• Docker Desktop
• OpenAI API key

Development

Run locally before deploying to AWS.

1. Clone the template

   git clone \
     https://github.com/agno-agi/agentos-aws-template.git \
     agentos-aws
   
   cd agentos-aws
   

   The template includes the following agents:

   Agent	Description
   Knowledge Agent	Answers questions from your documents using Agentic RAG.
   MCP Agent	Connects to external tools via Model Context Protocol.

2. Set your API key

   cp example.env .env
   

   Edit .env and add your OPENAI_API_KEY.

3. Start AgentOS

   ./scripts/venv_setup.sh && source .venv/bin/activate
   ag infra up --env dev
   
4. Load knowledge

   docker exec -it agentos-aws-template-api python -m agents.knowledge_agent
   

   This loads the default documents into the Knowledge Agent’s vector database.

5. Confirm it's running

   Navigate to localhost:8000/docs to see your AgentOS API.

6. Connect to the control plane

   1. Go to os.agno.com
   2. Click Connect OS → Select Local
   3. Enter http://localhost:8000

Your AgentOS is running locally.

Production

Deploy to AWS when you’re ready to go live.

Prerequisites

• AWS account
• AWS CLI installed and configured
• VPC with at least two subnets in different availability zones
• Container registry (Docker Hub or ECR)
• Session Manager plugin (for ECS Exec)

1. Verify AWS credentials

   aws sts get-caller-identity
   
2. Configure AWS settings

   Update infra/settings.py with your subnets and image registry:

   infra/settings.py

   infra_settings = InfraSettings(
       aws_region="us-east-1",
       aws_az1="us-east-1a",
       aws_az2="us-east-1b",
       aws_subnet_ids=["subnet-xxx", "subnet-yyy"],
       image_repo="your-dockerhub-username",
       push_images=True,
   )
   

   Set aws_az1 and aws_az2 to match your subnet availability zones. For ECR, use image_repo="[ACCOUNT_ID].dkr.ecr.[REGION].amazonaws.com".

3. Authenticate to your registry

   For Docker Hub:

   docker login
   

   For ECR, create the repository and authenticate:

   aws ecr create-repository --repository-name agentos-aws-template --region us-east-1
   aws ecr get-login-password --region us-east-1 | \
     docker login --username AWS --password-stdin [ACCOUNT_ID].dkr.ecr.us-east-1.amazonaws.com
   
4. Set production secrets

   cp infra/example_secrets/prd_api_secrets.yml infra/secrets/prd_api_secrets.yml
   cp infra/example_secrets/prd_db_secrets.yml infra/secrets/prd_db_secrets.yml
   

   Edit prd_api_secrets.yml and add your OPENAI_API_KEY. Edit prd_db_secrets.yml and set MASTER_USERNAME and MASTER_USER_PASSWORD for your production database.

5. Deploy

   ag infra up --env prd
   

   Takes about 5 minutes. The command builds and pushes your image, provisions RDS PostgreSQL, creates ECS resources, and configures an application load balancer.

6. Load knowledge

   Use ECS Exec to load documents into the Knowledge Agent’s vector database:

   ECS_CLUSTER=agentos-aws-template-prd
   TASK_ARN=$(aws ecs list-tasks --cluster $ECS_CLUSTER --query "taskArns[0]" --output text)
   
   aws ecs execute-command --cluster $ECS_CLUSTER \
     --task $TASK_ARN \
     --container agentos-aws-template-prd-api \
     --interactive \
     --command "python -m agents.knowledge_agent"
   
7. Get your endpoint

   Find your Load Balancer DNS in the AWS Console (EC2 → Load Balancers). Navigate to <load-balancer-dns>/docs to confirm the API is running.

8. Connect to the control plane

   1. Go to os.agno.com
   2. Click Connect OS → Select Live
   3. Paste your AWS endpoint

Your AgentOS is live on AWS.

Next Steps

Add apps

Browse agents, teams, and workflows for your deployment.

Connect interfaces

Expose your agents on Slack, Discord, WhatsApp, or MCP.