Velda is a cloud-native development, workload orchestration & HPC (High Performance Computing) platform. Directly scale your application from your development environment, no extra setup required.
- Seamless development experience - Connect with your favorite IDE (SSH, VS Code, Cursor, Windsurf), or code directly from your browser
- Instant onboarding - Clone from pre-configured images or customize your own
- Universal compatibility - Works with most libraries, tools, and package managers
- Persistent environments - All customizations are saved and isolated per user
- Simple scaling - Just prefix
vrun
to your command and run with requested resources - Run anything - ML training, batch processing, microservice clusters
- Unbounded capacity - Access as many machines as you need from your cloud provider
- Consistent environment - Your code, data, dependencies always mounted on new machines
- No idle resources - Only pay for what you use, no more idle GPUs or machines
- Reduced engineering overhead - Skip container image management
- Optional K8s-free scaling - Scale directly with VMs from your cloud provider
# Initialize mini-velda sandbox
velda mini init sandbox
# Connect to sandbox
ssh mini-velda
# Setup environment
sudo apt install python3
pip install torch
# Run workload with GPU
vrun -P aws:g6.xlarge train.sh
We love contributions and community engagement! Here's how to get involved:
Check out our main repository velda-io/velda
for detailed documentation, setup guides, and contribution guidelines.
Ready to scale your development workflow? Get started with Velda today!