Jarrod E. Brown
I build enterprise security solutions and systems from the ground up — and I'm now building the agentic AI meant to run on top of them.
About
My career has followed one consistent pattern: I'm handed something fragile or unbuilt, and I turn it into something that holds. I went straight to work out of high school and got into IT, starting on help desk, around 2000, and climbing up from entry level roles to data-center operations by 2009 — earning my degrees along the way: an associate's first in 2008, the bachelor's later in 2010, and the MBA in 2026. Nobody handed me the ladder; I built it one rung at a time.
Along the way I served in the U.S. Army Reserve as a 25B Information Technology Specialist — installing and securing field network infrastructure, handling COMSEC, and operating across classified and unclassified environments. The Army taught me that infrastructure is only as good as the discipline behind it; I finished Advanced Individual Training second in my class, within a tenth of a percent of the top, and I've carried that standard into everything since. I was honorably discharged as a Specialist (E-4).
I learned large-scale engineering inside a national higher-education enterprise, helping design and deploy a Cisco Nexus data-center fabric across three sites and a private MPLS-BGP backbone connecting data centers and campuses nationwide — enforcing Dev/QA and production separation with VRFs and validating designs in GNS3 before they ever touched production. That's where the architecture instinct took hold.
Moving into a CyberSecurity role at a global semiconductor manufacturer, I reported to the CISO and directed a worldwide migration from Cisco ASA, Fortinet FortiGate and Check Point firewalls to Palo Alto Networks — implementing Panorama, GlobalProtect for VPN, and multiple HA pairs of PAN-NGFWs while developing the Security Profiles for deep packet inspection. I designed the WAN modernization strategy that replaced managed MPLS with SD-WAN across U.S., EMEA, and APAC, and built the internet-edge and DMZ architecture for global co-location data-center builds.
Then I founded my own network-security consulting practice, delivering through VAR partnerships for financial-services, healthcare, and technology clients. I led zero-downtime migrations off legacy firewall platforms, integrated security into Cisco ACI fabrics for policy-driven micro-segmentation, and aligned client environments to CIS Benchmarks, HIPAA, and PCI-DSS. Running my own shop taught me the business behind the architecture — scoping work, owning outcomes, and earning trust before touching production.
That work carried me deep into financial services. At one bank I re-engineered the Palo Alto Panorama architecture, deployed AlgoSec for automated policy recertification, and administered Zscaler ZIA/ZPA for cloud-delivered secure access. At another I did the work I'm proudest of: standing up an enterprise Network Security team from nothing. I built the operating model and three service domains around it, drove enterprise Zero Trust integrating Arista, Palo Alto, and AWS/Azure controls, governed DMZ architecture across cloud and on-prem, cut firewall policy delivery from thirty days to three through automation, and found $500K in annual savings by right-sizing our micro-segmentation footprint — all while delivering security workstreams for major M&A integrations.
The MBA paired the architecture with the business case — coursework in data-driven decision making, IT strategy, and ethical leadership. Out of fifteen years of evolving systems that can't be shut down, I've been drawn to a methodology called WISER — a way for teams to innovate continuously with AI by rebuilding systems while they run rather than stopping to start over: action over theory, evolution over disruption, people over proxies. It already existed; I'm learning it and putting it into practice across the work below.
I now put that methodology to work building agentic AI. ServiceBay AI is a five-agent assistant on IBM watsonx Orchestrate that turns a vague dashboard warning into a verified diagnosis, recall check, cost estimate, and shop list, and an MCP server wires those agents into Claude. The largest thing I'm building is Calaite, a transcript-driven audio and video editor on Cloudflare that started as tooling for the podcast and is now a product under Isobaric Labs. Alongside them I keep building the security automation I know cold — a phased PAN-OS configuration refactor, and an EdgeRouter monitoring loop where cloud analysis of the network's own traffic feeds signed block lists back to the edge as enforcement. Each is documented on the projects page as a full solution architecture.
I also co-host The Broken Handle, a weekly podcast on the modern job market — layoffs, AI displacement, and career pivots — with my co-host William Lowdermilk. We're both veterans, and we built a Cloudflare-native research pipeline that ingests dozens of forward-looking labor indicators so our commentary is grounded in data rather than anecdote — plus the editor we use to cut the show, which turned into Calaite. The same forces I design around at work are reshaping careers everywhere, and the show is our honest read on them. This site is where I keep the record of what I'm building.