Cloud + Kubernetes AI Deployment Platform
Deployment-platform project that runs the completed black-box distillation API behind an observability gateway, adds Docker Compose and Kubernetes manifests, readiness/liveness probes, CI validation, Prometheus metrics, Grafana dashboards, and a cloud-ready GKE path.
Send a prediction through the Kubernetes-ready gateway and inspect workload health, metrics, and deployment targets.
Page expectations
What this route loads and what you should see.
Loads gateway status, upstream health, metrics contract, and deployment targets.
Returns the black-box student prediction through the platform gateway while recording request and prediction metrics.
Kubernetes-ready prediction
Send a prediction through the Kubernetes-ready gateway and inspect workload health, metrics, and deployment targets.
Returns the black-box student prediction through the platform gateway while recording request and prediction metrics.
Gateway status
Loads gateway status, upstream health, metrics contract, and deployment targets.
Platform prediction
Returns the black-box student prediction through the platform gateway while recording request and prediction metrics.
Grafana dashboard
Rendered preview of the Grafana dashboard provisioned with the deployed platform.
Cloud Kubernetes AI Platform
Request rate by status
Separates gateway request volume by HTTP status code.
ai_platform_requests_totalp95 gateway latency
Tracks the 95th percentile gateway latency over the last five minutes.
ai_platform_request_latency_seconds_bucketUpstream error rate
Highlights timeout, invalid-response, and upstream failure modes.
ai_platform_upstream_errors_totalPrediction distribution
- AI Engineer
- Data Scientist
- Sr. Data Scientist
Shows which student labels the live prediction path is returning.
ai_platform_predictions_totalThe live Grafana UI stays internal. This rendered preview is based on the committed dashboard definition at monitoring/grafana/dashboards/cloud-kubernetes-ai-platform.json.