Building the pipeline where humans set direction and AI handles the rest.

Splunk/Cribl consultant by day. Automating myself out of a job by night.

The goal: file a GitHub Issue, grab coffee, come back to a PR that's been implemented, tested, and reviewed by multiple AI models — just waiting for a thumbs up. Not fully there yet, but close enough to be dangerous.

Humans decide what to build. AI agents handle the how. Automation runs the boring parts. A human gives the final sign-off. Claude, Gemini, Copilot, and local MLX models each do what they're best at — the right model for the right job instead of throwing everything at one.

flowchart LR
    subgraph Human["Human"]
        direction TB
        H1["Roadmap"]
        H2["GitHub Issues"]
        H3["PR Review"]
    end
    subgraph AI["AI Agents"]
        direction TB
        A1["Claude / Gemini / Copilot"]
        A2["Code"]
        A3["AI Code Review"]
    end
    subgraph Auto["Automation"]
        direction TB
        T1["CI / Testing"]
        T2["Lint & Validate"]
        T3["Ship It"]
    end
    H1 --> H2 --> A1 --> A2 --> T1 --> T2 --> A3 --> H3 --> T3

On the Clock: Splunk and Cribl consultant specializing in security operations. I architect SIEM platforms, build detection pipelines, and optimize data flows. My specialty? Cutting ingest volume by 30-50% while actually improving security posture.

Outside the Terminal: When I'm not wiring up AI agents or debugging data pipelines, I'm probably over-engineering my home lab or convincing my fish that uptime matters.

Home Lab: Proxmox cluster, UniFi networking, Home Assistant, Splunk, Cribl — all managed with Terraform, Ansible, and Nix. The goal is fault-tolerant infrastructure I can rebuild from a single nix build.

Aquariums: 75-gallon saltwater reef (clownfish, corals, pistol shrimp) + freshwater tanks with custom lighting and wave-maker automations. The fish have better SLOs than most production systems.

Adventures: Scuba diving (San Pedro, Belize is my happy place), snowboarding in Michigan and Colorado, hiking, running.

AI Development Pipeline — I point Claude, Gemini, and Copilot at GitHub Issues and they figure out the rest. Multi-model routing with local MLX fallback for when I don't want to burn cloud tokens on a typo fix. ai-assistant-instructions ・ claude-code-plugins ・ ai-workflows

AI Observability — OTEL pipeline tracking every AI coding interaction from IDE to Splunk. If an AI agent touches code, I know about it — usage, cost, performance, all of it. VisiCore_App_for_AI_Observability ・ cc-edge-copilot-otel ・ cc-edge-vscode-io

Nix Reproducible Everything — Four repos, one goal: nix build and walk away. nix-darwin for macOS, nix-ai for AI tooling, nix-home for dev environment, nix-devenv for project shells.

Local LLM Inference — Running MLX-native models on Apple Silicon because why pay for tokens when you have 128GB of unified memory? Currently on M4 MacBook Pro, eyeing the Mac Studio. nix-ai (MLX module)

RAG & Context Engineering — Qdrant on the home lab feeding context into AI workflows. The goal: AI that actually knows your codebase instead of hallucinating import paths.

Home Lab IaC — Proxmox cluster, 30+ repos, everything in code. Terraform provisions, Ansible configures, Renovate keeps deps fresh, release-please handles versions. terraform-proxmox ・ ansible-proxmox ・ .github ・ secrets-sync