Docs
Cloud deployment
Cloud deployment
Run AITracer on common cloud providers using standard patterns.
Cloud deployment is a common choice for teams that want managed databases, elastic capacity, and familiar operational tooling.
This guide stays at the pattern level: what to provision and how it usually fits together. It does not document any vendor-specific internal architecture.
Typical shape
Rendering diagram...
You provide DNS, HTTPS, secrets, and observability using your cloud’s services or equivalents.
Providers
AITracer runs the same way on major clouds, for example:
- Amazon Web Services
- Microsoft Azure
- Google Cloud Platform
Choose regions and compliance offerings that match your requirements.
What to provision
Most teams allocate:
- a stateless application tier (containers or VMs) behind a load balancer
- managed PostgreSQL (or a self-managed cluster you support)
- logging and metrics from your cloud’s observability stack
Object storage is optional and depends on how you handle exports and long-term retention.
Deployment workflow
- Provision infrastructure with your IaC tool (Terraform, Pulumi, Bicep, CloudFormation, or the provider console).
- Configure environment variables and secrets (see Environment configuration).
- Run database migrations as part of your release process.
- Roll out the application with your standard blue/green, rolling, or canary strategy.
- Verify HTTPS, health checks, and ingest from a non-production workspace before promoting.
Best fit
Cloud deployment suits:
- growing teams that want elastic capacity
- SaaS products embedding AI observability
- enterprises standardizing on a single cloud provider
For sizing detail, see Infrastructure requirements.