Post by Adrian Tineo

Everyone is an "AI Engineer" these days. But are they? New certifications are bringing clarity to the scene. This is how I got my Claude Architect certification. The CCA-F (Claude Certified Architect - Foundations) exam grants you access to this new certification. It evaluates a wide range of competencies across various domains and scenarios. Including agentic architecture, tool design, Claude Code configuration, context management and more, this exam puts you to the test on really practical skills, which are crucial to project success. The scenarios are the part that really stood out for me, because it means that questions are not just abstract and academic, but rather concrete and hands-on. From customer support, to code generation, developer productivity, and multi-agent research systems, to name a few. Currently, you can only access this certification through approved partners, not individually (thanks to A.Team for the sponsorship!). If you are going for the exam, the authoritative document is the exam guide. Study it thoroughly because it closely matches the skills and domains you will be tested on. If you are active in the field, it shouldn't take too much preparation: I got 87% mark with only a couple of days of serious study. However, you've only got one attempt and passing mark is 72% so don't take it lightly. Although the exam is organized around Claude, via API, SDK or CLI, its underlying principles are universal to generative AI, so the principles are easily transferable to other models, including open source ones. Having said that, the exam does expect familiarity with Anthropic's specifics. Beyond the certification itself, it's encouraging to see best practices solidifying, coming together as a coherent body of work. These foundations are the set of practices for building the systems of the future.