I’m the founder of Amendly Labs, an independent research lab focused on a central question for the AI era: how do individuals and organizations preserve economic leverage as AI compresses the value of traditional expertise?
My work spans offline autonomous agents, long-horizon model evaluation, sovereign data tools, and human-centered systems designed for a future where intelligence is abundant but strategic control is not.
Rather than optimizing only for immediate task completion, I focus on qualities that are often under-explored in current evaluations: breadth of exploration, novelty generation, persistence, behavioral drift, and long-run reasoning patterns.
Current areas of work include:
- Offline memory-centric agent systems for long duration autonomous experimentation.
- Benchmark design for novelty, breadth, and sustained reasoning.
- Structured experiment traces for model behavior analysis.
- private offline tools that help individuals optimize work and behavior without exporting sensitive data.