An ongoing bilingual experiment
An AI-native organization is not a traditional company with Copilot added. It is designed around AI agents, shared memory, parallel execution, and human oversight. Ingora, a Singapore company founded in 2026, is testing through its Shanghai subsidiary whether five humans can cover capabilities that might once have required 40–50 people.
This is a capability-surface experiment—not a claim that five people equal fifty-person productivity.
000–004
When the founder was away for nine days, could AI-generated material become new company knowledge?
What we learned: AI can keep sensing, reading, and organizing, but candidate knowledge needs human review before it becomes company memory for which the organization will take responsibility.
Read the full story on GitHubWhy can the company remain stuck after every individual gets faster?
What we learned: Faster local tasks do not make the whole company faster. Memory, permissions, handoffs, validation, and feedback must be redesigned together.
Read the full story on GitHubIf AI only makes existing work a little faster, what are we missing?
What we learned: The larger change is that people can cross role boundaries and test assumptions earlier, while professional judgment and responsibility remain essential.
Read the full story on GitHubWhat happens to serial handoffs when agents can keep acting?
What we learned: Shared memory and continuous processing within boundaries can compress cross-department waits. Speed still cannot remove professional review or final accountability.
Read the full story on GitHubWhen tokens, agents, and code are no longer scarce, what still constrains the company?
What we learned: The bottleneck is human experience that has not been systematized: what is worth doing, what good looks like, and how a lesson becomes reusable.
Read the full story on GitHubDirect answers
Do five people really equal fifty? Does AI-native mean no engineers? What are Human on the Loop and the Octopus Organization?
Ingora is a Singapore company experimenting with what it means to build an AI-native organization. At this early stage, access to AI talent and hardware supply chains matters, so many of the stories on this site take place inside its Shanghai subsidiary in China.
The experiment is led by Jacky Hsu. The outcome remains unknown; next, we will keep testing where humans should supervise agents operating at high speed and in parallel.