Linux / DevOps Engineer | Amateur Radio & Big Data Enthusiast | Exploring AI Agent (MCP) Integrations
I’m a long-time Linux user (since the Slackware days) and active supporter of Open Source Software.
- Ubuntu user since 2005 (5.04), Ubuntu Member since ~2013
- Launchpad Package Maintainer since 2010
- Daily driver: Ubuntu & Oracle Linux (enterprise), Pop!_OS for gaming, Alpine for containers
I’ve worked across Red Hat, Fedora, CentOS, Arch, Gentoo, and more. If it’s Unix-y, I’ve probably broken it and fixed it.
I am bridging Amateur Radio + AI workflows, moving beyond static logging toward a modular, "Grand Central Station" architecture for station management.
- Infrastructure as Code → Terraform & Ansible with OCI, Nomad, Docker/LXC
- Distributed Systems → Vault, Consul, HA setups
- Big Data & Propagation
- wspr-ai-lite — Lightweight DuckDB + Streamlit UI for multi-GB archive analysis.
Architect of the ki7mt-mcp-hub — a unified ecosystem of AI-agent micro-servers for Amateur Radio.
- adif-mcp (
com.adif-mcp.validator) — The authoritative validator for ADIF 3.1.6 log compliance. - wspr-mcp (
io.ki7mt.wspr.researcher) — Local AI access to gigabytes of propagation archives via DuckDB. - qrz-mcp (
io.ki7mt.qrz.lookup) — Specialized XML Callbook micro-service. - Station Control → Roadmap includes
io.ki7mt.rigfor Hamlib integration and real-time SFI/K-index monitoring.
- Language/Env: Python 3.12+, uv (for zero-config tool deployment)
- Engines: Ollama (Local LLMs), Qwen2.5-Coder (32B/7B)
- Data: DuckDB, ClickHouse, Apache Arrow, FastAPI/Uvicorn
This stack enables safe, contract-driven station automation, allowing AI agents to navigate logs and propagation data with high-fidelity precision.
- M3 AI Workstation: Mac Studio 96GB UDM ( Local LLM Engine )
- Linux AI Workstation: EPYC 7302 128GB, NVIDIA 5090
- Storage Server: 5950X TrueNAS
- Virtualization: Proxmox hypervisor for clustered VMs/containers
- Networking: US-XG-16 | USW-PRO-24 PoE | pfSense
- ki7mt-mcp-hub → The central repository for all KI7MT MCP servers.
- WSPR Analytics → Big Data exploration of WSPR spots using PySpark & Arrow.
- wspr-ai-lite → Portable DuckDB + Streamlit with MCP integration.
- Open for discussions on Linux, DevOps, Amateur Radio, or AI Agents
- Reach me on GitHub Discussions or via ki7mt.io