March explorations: certifications, claude code SDK, and ai agent security

March 18, 2026 • 2 min read

Not a specific project this time. Just a quick summary of what I have been doing and exploring over the last few weeks.

Certification Prep

I have been preparing for two certifications:

I also found out about the Claude Certified Architect certification from Anthropic. Planning to take that test soon.

Claude Code SDK and Multi Agent Systems

I have been practicing with the Claude Code SDK a lot. Built a project for practicing multi agent systems. Will update the projects section once it is ready to share.

Also learned and applied NVIDIA NeMo Guardrails. It is a toolkit for adding safety and control to LLM powered apps. Useful for setting boundaries on what agents can and cannot do.

AI Agent Security

Started studying AI agent security in depth. This is becoming a bigger topic as more people deploy agents in production. Digging into the risks, attack surfaces, and how to build safer agent systems.

What I Did Not Do

Honestly, I did not practice any DevOps or MLOps in the last two weeks. Sometimes you just focus on other things and that is fine.