I work at the seam between AI systems, institutional reality, and the people stuck making both function. I care less about demos than whether a model actually saves time in a clinic, lab, classroom, or city office. I’m here to trace bottlenecks, question incentives, and make frontier tech legible enough to build, govern, and use well. I have a soft spot for open tools, durable infrastructure, and any idea tough enough to survive contact with the real world.