Nelson A.

If you are looking for an experienced IT Director and systems architect who can help you build, integrate, and operate AI agents in real-world environments, I can help. I have over 22 years of hands-on experience in enterprise IT, advancing from the helpdesk to an IT Director role. Today, I focus on designing and building AI agent systems. This combination of deep infrastructure experience and modern AI engineering gives me a practical understanding of both the technology and the business problems it solves. ### AI Agents & Governance I design and build AI agents that interact with tools, systems, APIs, and enterprise data to perform real-world tasks. I am the creator and lead architect of SAFi, an open-source runtime governance engine for agentic AI. This framework governs agent reasoning and actions, maintaining a deterministic, auditable record of execution. My AI stack includes Python, FAISS, Model Context Protocol implementations, and integrations with major LLM providers like OpenAI, Anthropic, and Gemini. ### Microsoft Azure Extensive experience designing, deploying, and supporting infrastructure on Microsoft Azure. This includes cloud migrations, virtual machines, networking, identity management, and enterprise workloads. ### Enterprise IT & Infrastructure More than two decades of experience managing Active Directory, DNS, DHCP, networking, firewalls, Windows infrastructure, Linux, and enterprise systems. I specialize in diagnosing complex problems, writing PowerShell automation, and getting critical systems running smoothly. ### Microsoft 365 Experienced with Exchange Online, SharePoint, OneDrive, Microsoft 365 migrations, and the broader Microsoft cloud ecosystem. ### Linux & Web Infrastructure Expert knowledge in Linux system administration and full-stack containerization. I build and maintain environments using Docker, Docker Compose, MySQL, Nginx, and Apache. I also have working experience with the LAMP stack for popular CMS platforms like WordPress, along with JavaScript, CSS, HTML, and Python. My approach is simple: understand the problem, build the right solution, and make it work in the real world