Anthropic Claude Certified Architect – Professional (CCAR-P) Lessons
38 lessons, one per exam-guide subdomain, in the order the guide teaches them. Every claim is cited to the official documentation.
Domain 1: Solution Design & Architecture
6 lessons · 17% of the exam
1.1Translating business problems into Claude solutions
Subdomain 1.1: Translate business problems into Claude-based AI solutions
19 min read
1.2End-to-end agent architecture: the loop, tool execution and feedback
Subdomain 1.2: Design end-to-end architectures
19 min read
1.3Workflow, agent and augmented LLM patterns for Claude
Subdomain 1.3: Select appropriate architectural patterns
19 min read
1.4Multi-agent design: when to split and how to brief subagents
Subdomain 1.4: Design multi-agent systems and orchestration strategies
2 pages · 28 min read
1.5Task decomposition patterns: chaining, routing, parallelization and orchestrator-workers
Subdomain 1.5: Apply decomposition techniques for complex problem solving
20 min read
1.6Business value pillars for Claude solutions
Subdomain 1.6: Align solutions to business value pillars
19 min read
Domain 2: Claude Models, Prompting & Context Engineering
5 lessons · 13% of the exam
2.1Choosing a Claude model: lineup, starting point and cost per task
Subdomain 2.1: Select appropriate Claude models based on trade-offs
17 min read
2.2System Prompts and Reusable Prompt Templates for Claude
Subdomain 2.2: Design system prompts, templates, and guardrails
2 pages · 19 min read
2.3Few-shot and chain-of-thought prompting for Claude
Subdomain 2.3: Apply prompt engineering techniques
17 min read
2.4Context window budgeting: what counts and how to measure it
Subdomain 2.4: Optimize context windows and manage token usage
2 pages · 24 min read
2.5Prompt reuse with caching, modular prompts and Agent Skills
Subdomain 2.5: Implement prompt reuse strategies
20 min read
Domain 3: Integration
8 lessons · 19% of the exam
3.1Capability bloat in agent tool configurations
Subdomain 3.1: Evaluate tool/agent configuration for capability bloat
19 min read
3.2MCP authorization: discovery, client registration and scoped tokens
Subdomain 3.2: Analyze authentication and authorization requirements to identify security gaps
2 pages · 25 min read
3.3Accuracy-latency trade-offs: effort, fast mode, streaming and justified configs
Subdomain 3.3: Evaluate accuracy-latency trade-offs and justify configuration decisions
17 min read
3.4Run-level telemetry and tagging for Claude agent fleets
Subdomain 3.4: Analyze observability challenges and select monitoring strategies at scale
2 pages · 25 min read
3.5RAG ingestion: chunking, embeddings and contextual indexing
Subdomain 3.5: Design a RAG pipeline with appropriate chunking and indexing strategies
2 pages · 22 min read
3.6Retrieval strategies matched to data shape and query pattern
Subdomain 3.6: Apply retrieval strategies matched to data shape and query pattern
18 min read
3.7Choosing between MCP, direct tools and agent-to-agent integration
Subdomain 3.7: Evaluate connection protocols and select the appropriate integration mechanism
22 min read
3.8Progressive discovery vs. monolithic context: tool search and when to use it
Subdomain 3.8: Evaluate progressive discovery vs. monolithic context strategy
17 min read
Domain 4: Evaluation, Testing & Optimization
6 lessons · 16% of the exam
4.1Evaluation metrics for Claude systems: accuracy, latency, cost, safety and security
Subdomain 4.1: Define evaluation metrics
20 min read
4.2Eval datasets and graders: designing a mixed-method test framework
Subdomain 4.2: Design evaluation datasets and test frameworks using mixed methodologies
18 min read
4.3A/B testing prompts: baselines, segments and versioned rollout
Subdomain 4.3: Conduct A/B testing and iterative improvements
18 min read
4.4Diagnosing Claude system failures: hallucination, prompt, tool and model faults
Subdomain 4.4: Diagnose system issues
17 min read
4.5Optimizing Claude cost and latency: free wins, batching and budgets
Subdomain 4.5: Optimize token usage, latency, and cost-performance trade-offs
22 min read
4.6OpenTelemetry for Claude Code and the Agent SDK: signals, traces and content gates
Subdomain 4.6: Monitor system performance using logging and observability tools
2 pages · 25 min read
Domain 5: Governance, Safety & Risk Management
5 lessons · 14% of the exam
5.1Input and output guardrails for Claude applications
Subdomain 5.1: Implement guardrails and safety controls
2 pages · 29 min read
5.2LLM failure modes: hallucination, knowledge limits, context rot and prompt injection
Subdomain 5.2: Identify risks, limitations, and failure modes of LLM systems
20 min read
5.3Human approval checkpoints in Claude Code and the Agent SDK
Subdomain 5.3: Apply human-in-the-loop validation strategies
2 pages · 25 min read
5.4HIPAA, BAAs and zero data retention for Claude
Subdomain 5.4: Ensure compliance with regulations
2 pages · 26 min read
5.5Bias, fairness and honest outputs in Claude builds
Subdomain 5.5: Address ethical AI considerations
2 pages · 18 min read
Domain 6: Stakeholder Communication & Lifecycle Management
5 lessons · 14% of the exam
6.1Structured discovery and requirement gathering for Claude solutions
Subdomain 6.1: Conduct structured discovery and requirement gathering
19 min read
6.2Communicating architecture trade-offs to stakeholders
Subdomain 6.2: Communicate architectural decisions and trade-offs
21 min read
6.3Stakeholder expectations: SLAs, limits and pilot feedback
Subdomain 6.3: Manage stakeholder feedback loops and expectation alignment
19 min read
6.4Architecture documentation for Claude solutions
Subdomain 6.4: Document architectures and provide implementation guidance
2 pages · 22 min read
6.5AI solution lifecycle phases: discovery to retirement
Subdomain 6.5: Support lifecycle phases
19 min read
Domain 7: Developer Productivity & Operational Enablement
3 lessons · 7% of the exam
7.1Claude Code settings scopes, precedence and permission rules
Subdomain 7.1: Configure Claude tools and environments for teams
2 pages · 24 min read
7.2Explore, plan, implement, commit: the Claude Code loop
Subdomain 7.2: Improve developer workflows using AI-assisted tooling
2 pages · 28 min read
7.3Diagnose Claude Code configuration: what loaded and why
Subdomain 7.3: Support debugging and operational issue resolution
2 pages · 23 min read