3amon_2
Remote: Yes. Remote, hybrid, or onsite around the Bay Area are all fine.
Willing to relocate: No
Technologies: Python, C/C++, TypeScript. Zephyr, FreeRTOS, nRF52840, ESP32, Thread, BLE, LoRa, signed OTA, KiCad, some VHDL. MCP servers, agent frameworks, LLM-judged evals, Claude Code. Kubernetes/GKE, Terraform, Cloudflare Workers, OpenTelemetry, Grafana.
Résumé/CV: https://3amon.net/Eamon-White-Resume.pdf
Email: [email protected]
Years of experience: 19
I'm an embedded engineer who spent the last year building AI agents, and I'd like to keep doing both. Most of my career was firmware and real-time work: optical terminals for satellites (bare-metal Cortex-M7, test campaigns at the Naval Research Lab), FPGA and driver work at National Instruments, Linux point-of-sale hardware. This past year I built an SRE on-call agent for an enterprise AI platform. It reads the production alert queue and investigates through policy-gated MCP tools on GKE. I also built the MCP tool router it runs on.
Some things I've built on my own time, with write-ups at https://3amon.net and code at https://github.com/3amon:
- 39 wearable LED rings that keep a shared clock with no leader. nRF52840, Zephyr, OpenThread, signed OTA with rollback, my own carrier PCB, and simulators that gate every release.
- LoRaPeerLink, device-to-device LoRa without LoRaWAN. Built against a channel simulator, then 99.6% delivery over a one hour soak on real SX1262 radios.
- Guard rails for coding agents: fail-closed hooks, scoped MCP servers, and a little tool that shows every agent session running across my machines.
Looking for senior or staff roles in embedded/connectivity (device fleets, OTA, radios) or agent infrastructure (MCP, tool runtimes, agents that touch production). Best case is a team that needs both. Full-time preferred, contract is fine too.