Shea Duncan

Personal hub

On a mission to help deliver value that leads to abundance

Abundance gets built, not summoned.

Current focus

Three active threads

  1. Livit AI

    Reliability intelligence, experiential AI, and patent-pending GPU expansion hardware.

    Open

  2. Local LLM project

    Personal exploration of on-device inference and the tooling around it.

    Open

  3. Tinkering with building robots

    Leveraging 3D printing, 3D modeling, a Comma 4, and a Comma 3.

    Open

The arc

From reliability work to industrial AI

I spent a decade keeping critical infrastructure running: rotating equipment, functional safety, CMMS work, and the unglamorous discipline of systems that have to stay online.

Now the work is AI infrastructure for industry: products, on-prem inference, and hardware that respects the operating constraints of refineries, LNG terminals, and other places where the cloud is not always the answer.

Connect

Useful conversations only

Industrial AI deployments, local inference, PCIe/GPU hardware, and practical reliability problems are the right starting points.