Claude API production application
Tools, structured output, caching, evaluation, observability, and failure handling.
Evidence: Runnable service, evaluation report, trace sample, and production-readiness checklist.
CCA v2.11
Course completion, CCA app progress, and the official Claude Certified Architect credential are three different things. This page keeps them separate and shows the work still required before this platform can claim full exam readiness.
Academy
14 imported completions
One further confirmed completion awaits import
CCA content
141/168 detailed notes
27 outline-only lessons remain
Exam bank
15/150 questions
Current set is practice, not a mock exam
Official status
Access not yet confirmed
Partner Network application is the gate
Priority 1 · official access
Lead with the CCA platform, Claude architecture work, books, production apps, micro-consulting, and a concrete 24-minute coaching offer. Anthropic states that partner applicants gain Partner Academy access, which includes certification exams. Its June 2026 Services Track update also says new Registered applicants make a minimum commitment to 10 certified practitioners, so the application needs a credible team or associate pathway.
Open Partner NetworkModel Context Protocol: Advanced Topics
Finish course and replace 2 outlines
Claude Platform 101
Complete and import current platform baseline
Building with the Claude API
Replace 12 outlines with detailed notes
Claude with Google Cloud's Vertex AI
Finish course and replace 7 outlines
Claude with Amazon Bedrock
Finish course and replace 6 outlines
AI Fluency for Builders
Finish course and import curriculum
AI Fluency for Small Businesses
Import confirmed completion
Tools, structured output, caching, evaluation, observability, and failure handling.
Evidence: Runnable service, evaluation report, trace sample, and production-readiness checklist.
Server and client with tools, resources, prompts, authentication, and notifications.
Evidence: Working integration, protocol trace, threat model, and transport decision record.
CLAUDE.md, Skills, subagents, hooks, GitHub automation, and the SDK.
Evidence: Reference repository, automated workflow, safety boundaries, and maintainers' runbook.
Direct Anthropic API versus AWS Bedrock versus Google Vertex AI.
Evidence: Decision matrix covering cost, governance, portability, operations, and deployment patterns.
Validate the schema, source links, difficulty labels, distractor rationale, and weak-area tracking before scaling to 150.