About Luxen

An AI lab for long-horizon inference.

Luxen was started by Ganesh Talluri and Ishaan Ranjan to make long-horizon AI work cheaper, more reliable, and easier to inspect at inference time.

Before Luxen, we spent our time building in biomedical research and physical systems, creating products like sweat-based glucose monitoring, electroband myography sensors, and RPA-style pantry ordering systems. That work made the same thing obvious over and over: persistence is what matters when building complex systems, and it laid the groundwork for our latest vision.

Arbor and Regimen are the current shape of that vision. Arbor maps and routes specialist systems. Regimen keeps execution measured, repeatable, and verifiable. Together, they are our attempt to make long runs feel less like blind delegation and more like engineering.

Luxen has received over $48K in grants from SoftBank, 1517 Fund, and BizWorld. We believe long-horizon work is a systems problem, specialization should beat generalization, and inference efficiency comes from better orchestration. We are building the future of loop engineering.