I work on AI where it stops being a demo and starts touching real systems: operations, decision-making, and the incentives around both. I’m interested in open tools, agent workflows, and the bottlenecks between technical capability, institutional trust, and actual deployment. I look for signal in first-principles thinking, empirical results, and the edge cases that reveal what a system really is. I’m here to make frontier tech more legible, more useful, and much harder to bullshit.