Claude Code Certified Architect Foundation (CCAR-F): Certification Bootcamp
Building AI agents with Claude is becoming easier. Designing systems that are reliable, scalable, and ready for production requires a different set of skills.
From choosing the right agent architecture and designing effective tool integrations to managing context, enforcing safeguards, and handling failures, every architectural decision affects how well an AI system performs.
Anthropic's Claude Certified Architect – Foundations (CCAR-F) certification tests your ability to make these decisions across real-world scenarios. Knowing the concepts is only part of the challenge. You also need to understand when to apply them, evaluate different approaches, and identify the most appropriate solution.
This 2-day, 10-hour live certification bootcamp, led by Puria Izady, is designed to help you build that understanding and prepare for the CCAR-F exam.
Through 6 runnable labs, scenario-based exercises, exam-style questions, and a guided simulated exam, you'll work across all five certification domains, including Agentic Architecture, Tool Design and MCP, Claude Code, Prompt Engineering, and Context Management and Reliability.
You'll also receive a full-length 60-question practice exam and supporting resources to continue preparing after the workshop.
What You'll Learn
Day 1: Agentic Architecture, Tool Design & MCP
November 21 | 5 Hours
Explore how Claude-powered agents interact with tools, coordinate tasks, and enforce reliable workflows.
Tool Design and Selection: Design tool definitions, descriptions, and schemas that help Claude choose the right tool for a task.
Model Context Protocol (MCP): Understand MCP clients, servers, resources, tools, configuration, and structured error handling.
Agentic Architecture: Work with agent loops, the Claude Agent SDK, tool permissions, and execution control.
Multi-Agent Orchestration: Design coordinator-subagent workflows, delegate tasks, manage parallel execution, and combine results.
Safety and Enforcement: Use hooks, policy gates, and human escalation to enforce rules beyond prompting.
Exam-Style Practice: Work through customer support and multi-agent research scenarios, including explanations of correct and incorrect answers.
Day 2: Claude Code, Prompt Engineering & Reliability
November 22 | 5 Hours
Learn how to configure Claude Code for engineering teams, produce reliable structured outputs, and manage complex AI workflows.
Claude Code Configuration: Work with CLAUDE.md, project rules, custom commands, skills, and team-level configurations.
Development Workflows and CI/CD: Use planning modes, headless execution, automated code review, and testing workflows.
Prompt Engineering: Apply precise instructions and few-shot examples to improve consistency and reduce false positives.
Structured Outputs: Design schemas, validate extracted information, manage retries, and handle ambiguous data.
Context Management: Manage context budgets, compaction, sessions, recovery, and large-codebase exploration.
Reliability and Human Review: Handle partial failures, conflicting information, confidence calibration, and escalation.
Simulated Certification Exam: Work through a timed, scenario-based practice exercise covering all five certification domains.
What Can You Expect?
This isn't just a series of lectures covering certification terminology. The workshop is structured around the decisions you'll need to make in real-world Claude applications and scenario-based certification questions.
You'll get:
10 hours of live instruction across two days
6 runnable labs covering MCP, multi-agent workflows, policy enforcement, code review, structured extraction, and context management
Exam-style questions integrated throughout both days
Scenario-based exercises across all six exam settings
A guided simulated exam with explanations of why each answer is correct or incorrect
A full-length, 60-question practice exam for independent revision
What You'll Walk Away With
By the end of the bootcamp, you'll have:
A structured understanding of all five CCAR-F certification domains
A decision checklist covering the exam's 30 task statements
Practical experience applying architectural concepts through six runnable labs
Experience answering and reviewing certification-style questions
A full-length practice exam to help identify areas requiring further revision
Reusable lab notebooks and supporting resources for continued practice
Meet Your Instructor
Puria is the creator of Pumping Code, an engineering education platform with more than 10,000 students across 110 countries.
Previously a Solutions Architect at Amazon Web Services, he worked with software companies across EMEA and helped more than 40 enterprises bring generative AI systems into production. His experience includes supporting launches of Amazon Bedrock, AgentCore, and the Strands Agents SDK in Germany.
He brings practical experience in cloud architecture, production AI systems, and developer education to the bootcamp.
Who Is This Workshop For?
This bootcamp is designed for:
Solution Architects designing Claude-powered applications and agentic systems
AI Engineers building production-ready AI agents, tool integrations, and multi-agent workflows
Software Developers working with Claude Code, APIs, and AI-assisted development
Senior Engineers and Tech Leads responsible for architecture, reliability, and technical decisions
CCAR-F Certification Candidates looking for structured, exam-focused training and practical exercises
Level: Intermediate
You'll benefit most if you have programming experience, familiarity with APIs, and a basic understanding of LLMs or AI agent workflows.