Architecture Lab for Cloud-Native Depth
I recently shared why I’m stepping off the whiteboard and back into the terminal. Today I’m introducing how that journey becomes real: a public project called Bugle Call.
What is Bugle Call?
Bugle Call is a purpose-built architecture lab.
It mimics the concerns platform teams handle every day:
- Incident ingestion under burst and idle load
- Asynchronous processing with enrichment and ML tagging
- Secure dashboards for incident trends and cost insights
- Observability that measures health and spending
- GitOps delivery across dev, staging, and production
- Multi-cloud deployment on kind, AKS, and EKS
The stack mixes .NET, Go, Python, React, Kubernetes, Terraform, Prometheus, and Kubecost.
Raw payloads land in MongoDB for audit and model training. Structured data lives in PostgreSQL for queries.
Why build it?
- Depth over theory – real code and live clusters reveal trade-offs that diagrams hide.
- Breadth with purpose – every component teaches a specific skill: IAM, GitOps, cost telemetry, or AI enrichment.
- Public proof-of-work – architecture credibility grows when anyone can clone, run, and break your design.
Repo and docs
| Resource | Link |
|---|---|
| Main repository | https://github.com/AdrianFreemantle/bugle-call |
| Project README | https://github.com/AdrianFreemantle/bugle-call/blob/main/README.md |
| Architecture doc with diagram | https://github.com/AdrianFreemantle/bugle-call/blob/main/docs/architecture.md |
Next steps
- Flesh out the Incident API and deploy to a local kind cluster
- Wire GitHub Actions for container build and linting
- Push manifests through ArgoCD into a dev namespace
- Write about early GitOps pitfalls and how autoscaling behaves on day one
Follow progress on the repo and join the conversation. If you are exploring similar territory, open an issue or drop a comment. Let’s compare notes while we build systems that can take a punch.
Top comments (0)