Challenge catalog
Start with free guided labs. Each challenge has objectives, commands, expected signals, paste-output validation, progressive hints, and solution reveal tracking.
Showing 12 of 12 challenges
Deploy a GPU-backed OpenAI-compatible endpoint and prove scheduling, health, TTFT, queueing, and rollback readiness.
Operate ingestion, metadata filters, vector retrieval, answer evaluation, and failure drills for production RAG.
Run a launch review across security, quota, rollout, observability, cost, and ownership before live traffic.
Build the signal model needed to debug user latency, runtime saturation, GPU pressure, traces, logs, and alerts.
Design the deployment contract for vLLM with model cache, readiness, runtime flags, and service exposure.
Choose the serving abstraction by ownership model, CRDs, graph complexity, autoscaling, and rollout needs.
Prove accelerator placement with labels, taints, tolerations, quotas, and unschedulable-pod debugging.
Measure retrieval recall, citation accuracy, tenant filtering, and reranking latency before generation.
Calculate cost per request from input tokens, output tokens, GPU profile, utilization, and cache behavior.
Design traffic shifting, readiness gates, rollback triggers, and model-version ownership for inference services.
Review tenant routing, namespace boundaries, secrets, NetworkPolicy, prompt logging, and retrieval authorization.
Create a dashboard model that joins user latency, queue wait, GPU pressure, token throughput, and cost signals.