Building an AI Operating System

Everyone wants an AI native operating model. Few have taken one from a handful of experiments to something a whole company runs on.

Felix Geelhaar did it at PandaDoc. He walks through the phases of the rollout, how teams were actually enabled rather than just handed tools, and the guardrails that kept the new speed from breaking an established product.

Then the impact, which nobody plans for. Engineering stopped being the bottleneck. Product caught up and stopped being the bottleneck. Now it is go to market, and behind go to market sits the customer. You will leave with the phases, what each one demands of leadership, and the question waiting at the end of it. Once you have eliminated yourself as the bottleneck, what is a product person actually for?