Building intelligent infrastructure

“Enterprise infrastructure and AI aren’t separate disciplines — the best systems need both.” — John Murphy, NeuralConfig

The long way to AI

For more than 20 years, I've built and operated enterprise network infrastructure — routing protocols, wireless architectures, and zero-trust security — while also working across Linux, storage, virtualization, and cloud platforms. As a pre-sales Solutions Architect, I translate technical complexity into business outcomes. Today NeuralConfig builds production AI systems: live SaaS products, production Claude agents, and code-mode MCP servers that operate enterprise infrastructure safely.

I was building agentic AI systems before most people had a name for them — autonomous tools that SSH into switches, classify devices, resolve tickets. The gap was obvious: AI developers didn’t understand infrastructure, and infrastructure engineers hadn’t touched AI.

01

Enterprise Infrastructure

Enterprise networks across many verticals. More than 12 years in technical pre-sales, learning that production systems require more than technical specifications — they demand reliability, security, and operational excellence. CISSP, CCNP, KCNA, and multiple wireless certifications while specializing in enterprise wireless architecture and zero-trust security.

02

Adding AI to the Stack

After completing Stanford's Machine Learning specialization, I built my first AI agent for network automation using Claude. RAG systems, autonomous agents, and agentic workflows that solve real infrastructure challenges — combining deep infrastructure knowledge with AI capabilities.

03

Open Source & Community

Selected tools shared as open source on GitHub: autonomous agents for support automation, RAG-powered knowledge systems, and zero-touch provisioning platforms. Projects like ruckus-ztp and osticket-agent have gained community traction, validating demand at the intersection of infrastructure expertise and AI.

04

What's Next

Small models at the edge, multi-modal AI in production infrastructure, and agentic systems that operate across network, cloud, and security boundaries. The interesting problems are where AI meets real-world constraints — latency, reliability, and scale.

The full stack, both kinds

Deep expertise across AI/ML, enterprise networking, cloud infrastructure, and security. Not just theory — production systems running in real environments.

AI & Development
Python FastAPI Claude MCP Agentic AI OpenAI RAG Vector Databases LangChain Swift/iOS
Infrastructure & Security
CISSP Zero Trust BGP/OSPF/MPLS Enterprise Wi-Fi Wireless Architecture Linux Virtualization
Cloud Platforms
Cloudflare AWS GCP Azure Kubernetes Docker Terraform
Automation & Observability
Network Automation ZTP CI/CD Prometheus Grafana

Production AI systems

Selected production work that isn't public code — shown here as architecture, not repositories. These put AI to work in production: against real enterprise APIs and on live customer calls.

Code-mode MCP servers
A reusable code-mode MCP pattern exposing 1,500+ enterprise-network API operations to Claude through a single tool call — OAuth 2.1 with dynamic client registration and PKCE, on Cloudflare Workers. One architecture, applied across three vendor platforms.
MCP Claude Cloudflare
StrandCalls — AI voice agent
A production conversational-voice agent that answers business calls, books appointments into live calendars, and sends SMS follow-ups — natural speech, not IVR menus. Live SaaS on Cloudflare.
Voice AI Cloudflare SaaS

15+ projects on GitHub

Tools built to solve real problems across AI, networking, and security. All open source.

jira-analyzer
AI-powered pipeline for analyzing Jira feature requests. Automates extraction, Salesforce revenue enrichment, LLM-based vertical rating, and priority ranking.
Python AI Pipeline
osticket-agent
Autonomous AI agent for osTicket that uses NLP and tool-calling to analyze tickets, interact with network infrastructure, and generate intelligent responses.
Python Agentic AI
ruckus-ztp
AI-powered Zero-Touch Provisioning agent with a natural language chat interface for enterprise switches and access points.
Python Agentic AI Network
device-profiler
Automated network device profiling system that detects, fingerprints, classifies, and provisions network access for unknown devices.
Python Security Network
r1-sdk
Python SDK for a major enterprise Wi-Fi management platform — 161 operations across 18 modules with OAuth2 auth and PyPI distribution.
Python SDK
sz-acl
Advanced ACL management system for network security, featuring automated rule generation and conflict detection.
Python Security
pflogs
PF firewall log analysis with IP geolocation and threat pattern detection.
Python Security
wifi-test
Wi-Fi connection testing with custom MAC address support and signal diagnostics.
Network
wispr-portal
WISPr captive portal debugging and testing tool.
Network
r1-scripts
Automation scripts for a major enterprise Wi-Fi platform — bulk operations, reporting, and migration utilities.
Python Automation
openrouter-chat-app
AI chat interface leveraging OpenRouter's API to access multiple LLM providers through a single endpoint.
Python AI
survival-rag
Retrieval-Augmented Generation system demonstrating practical RAG implementation and vector database integration.
Python RAG

Certifications — current and previously held

Industry certifications across security, AI/ML, cloud, and networking.

Security & AI
Cloud & Networking
  • Kubernetes and Cloud Native Associate (CNCF)
  • CCNP (Cisco)
  • Certified Wireless Design Professional (CWDP)
  • Certified Wireless Security Professional (CWSP)
  • Ekahau Certified Survey Engineer

Reliable AI is responsible AI

AI systems that interact with production infrastructure must be built with safety as a first principle. Every autonomous agent includes guardrails, approval workflows, and comprehensive logging.

  • Human-in-the-loop for destructive or irreversible actions
  • Comprehensive audit logging for all AI decisions
  • Rate limiting and circuit breakers
  • Graceful degradation when facing uncertainty
  • Explicit approval workflows for production changes

The original portfolio

Where it all started — the interactive terminal demo that first represented NeuralConfig online.

root@neuralconfig:~

Let's build something

AI infrastructure, agentic systems, or the next hard problem.

Get in Touch View GitHub