CCAR-P
Claude Certified Architect (Professional)
Who it is for: Architects with 3+ years in systems or platform architecture.
63 questions · 120 minutes
27 lessons · 7 domains
Take a practice examHow do I take the exam?
- Your company must be part of the Claude Partner Network. Joining is free.
- On the Anthropic Partner Academy, open the certification page, read the guide and the terms, and register. The list price is USD 175; checkout shows the final price, with any discount for your company's partner tier.
- After confirming, create your Pearson VUE account and pick a date: online proctored or at a Pearson test centre.
- You can cancel or reschedule up to 24 hours before. After that, the fee is forfeited.
- The credential lasts 12 months. Renewing on time is free: a short, non-proctored assessment on what has changed.
Domains
Solution Design & Architecture 17%
- Workflow, agent or augmented LLM, the problem picks the pattern
- From "we want an agent" to the real problem
- Input, processing, output and feedback, what goes in code and what goes to the model
- Decompose and orchestrate, chaining, map-reduce and well-delegated subagents
Claude Models, Prompting & Context Engineering 13%
- Choose the model with evidence, not by size
- System prompts and templates that scale, sections, boundaries with their reason and untrusted content kept apart
- Zero-shot, few-shot or step-by-step reasoning, which to use depending on where it fails
- Context as a budget, prompt caching, modular prompts and Skills
Integration 19%
- Least privilege and authorization, removing capabilities beats watching them
- MCP, API/CLI or agent-to-agent, choosing the connection mechanism
- RAG pipeline, chunking by structure, contextual retrieval and version metadata
- Retrieval matched to the shape of the data, and accuracy versus latency
- Progressive tool discovery and agent observability at scale
Evaluation, Testing & Optimisation 16%
- What to measure and with which data, success criteria and mixed methods
- Comparing versions without fooling yourself, offline baseline, A/B tests and guardrail metrics
- Diagnosing failures, prompt, hallucination, retrieval or model mismatch
- Optimizing cost and latency and monitoring quality, not just uptime
Governance, Safety & Risk Management 14%
- Layered guardrails, the prompt is the first barrier and not the only one
- LLM failure modes, calculations, non-determinism, hallucination and the audit trail
- Human in the loop by risk, and how to avoid automation bias
- Regulation and fairness, where you deploy and how bias is measured
Stakeholder Communication & Lifecycle Management 14%
- Discovery and clashing requirements, who to bring in, what to write and who decides
- Expectations and SLAs, no 100%, targets against the baseline and latency by percentile
- Documenting to operate, runbook, ADRs and a reference implementation
- Handoff is not the end, an owner, scheduled evaluations and thresholds that trigger iteration
Developer Productivity & Operational Enablement 7%
- Claude Code for a team, what is shared, what is imposed and what is enforced
- AI-assisted development workflows and incidents, plan before implementing, reproducible CI and rate limits