I build automation systems for the messy part of operations: the work that sits between networking, security, APIs, AI, and the humans who still have to trust the result.
I've spent 15+ years across network engineering, security automation, solution architecture, and sales engineering. Today I'm a Principal Systems Engineer at Tines, working with teams on automation and AI-assisted operations.
A lot of my current work is focused on a question I think we're going to be dealing with for a long time:
How do we use AI in operations without giving up the engineering controls that made automation trustworthy in the first place?
That has me working on AI agents, MCP, RAG, tool use, structured outputs, validation, human approval, observability, and the guardrails needed before any of this gets near production infrastructure.
I'm also the author of Building AI Agents for Network Operations, where I take those ideas and apply them specifically to NetOps.
Most of what I build and write about falls somewhere around:
Network automation and NetOps
AI agents, MCP and RAG
Tines and workflow automation
SecOps and threat intelligence
Infoblox, APIs and Python
Automation architecture and governance
I also teach hands-on workshops because I learn far more from building something than staring at another architecture diagram, and I assume most engineers do too.
I publish the things I'm experimenting with, including the failures and weird edge cases, at
sifbaksh.com.