Docker Hybrid: Local + Cloud AI Agents

Why Go Hybrid?

Optimize for cost, speed, and scale

Cost Optimization

Run quick tasks locally (free). Scale intensive workloads to cloud only when needed. Pay for cloud resources only during active use.

Instant Local Feedback

Keep development agents local for zero-latency editing and testing. See changes immediately without network delays.

Unlimited Cloud Power

Offload memory-intensive builds, parallel tests, and data processing to cloud containers with 64+ GB RAM and GPU access.

Data Locality

Process large datasets in the cloud where they live. Avoid slow downloads—run agents directly near AWS S3, GCS, or database servers.

Auto-Scaling

Cloud containers scale based on workload. Spin up 50 test containers for parallel execution, then scale back to zero when done.

Unified Management

One dashboard for all agents—local and cloud. See resource usage, logs, and status for every container regardless of location.

How It Works

Four steps to hybrid architecture

1

Set Up Local Docker

Install Docker Desktop and AI Maestro on your local machine:

git clone https://github.com/\
23blocks-OS/ai-maestro.git
cd ai-maestro
./install.sh

2

Configure Cloud Workers

Add your cloud VMs or container services via Settings. Configure SSH keys for secure connections.

# In AI Maestro Dashboard
# Settings → Workers
# Add: cloud-vm.example.com:23000
# Upload SSH key

3

Deploy Strategically

Launch lightweight containers locally. Deploy resource-intensive containers to cloud workers:

# Local: Quick edits
docker run -d my-agent:latest

# Cloud: Heavy builds
ssh cloud docker run -d \
  --memory=32g my-build-agent

4

Unified Dashboard

AI Maestro shows all containers (local + cloud) in one interface. Monitor resources, view logs, switch seamlessly.

Perfect For

💻 Resource-Constrained Laptops

Working on a MacBook Air with 8GB RAM? Run quick edits locally, offload builds and tests to a cloud VM with 64GB RAM.

🧪 Large-Scale Testing

Test suite takes 2 hours locally? Spin up 20 cloud containers, run tests in parallel, complete in 10 minutes. Pay only for those 10 minutes.

📊 Data Science Workflows

Experiment with code locally. Train models on cloud GPUs. Process datasets where they live (AWS S3, BigQuery) without downloading terabytes.

💰 Cost-Conscious Teams

Maximize cloud ROI by running only compute-intensive tasks in the cloud. Use spot instances for 70% cost savings on non-critical workloads.

Get Hybrid Power
Local Speed + Cloud Scale

Set up hybrid architecture in 20 minutes. Works on macOS, Linux, and Windows (WSL2).

View Hybrid Setup Guide Compare Work Modes

Docker Hybrid Local Speed + Cloud Power