The Origin
OTP started with a question: if everything about how your AI team operates were lost, what single artifact would contain enough intelligence to rebuild it?
That question came from running 14 AI agents in production at Sneeze It, a digital agency. Agents named Radar (Chief of Staff), Dan (Strategic Co-Founder), Dash (Performance Analyst), Pepper (Executive Assistant), Crystal (Project Manager), Dirk (Revenue Operator), Neil (Chief Learning Officer), and more. Each with a clear seat, clear tools, and clear accountability.
The coordination rules that kept them from colliding lived in a single document. We called it the Organizational Operating System. Other organizations wanted to compare theirs, so we built OTP to share them.
Then a user asked, "can I add humans to the chart?" We added them. Then we put AI agents under those humans. Then we let humans author SOPs that the agents inherit at runtime. The product expanded -- the OOS was never only for agents. It was always for the whole organization. OTP today carries that forward: humans, AI agents, and the SOPs they share.
The Mission
Every organization is solving the same coordination problems from scratch -- between humans, between AI agents, and between the two. Who owns what? Which SOP governs which decision? What does an AI inherit when a human leaves? When does human oversight engage?
These answers exist, scattered across thousands of organizations. OTP makes them visible, comparable, and inheritable. The chart is the surface. The SOPs are the substance. The network is the leverage.
The Stack
The AI work-layer has three pieces. MCP and A2A are infrastructure between agents and their tools or each other. OTP is the layer above: the organization itself.
- MCP (Model Context Protocol): Agent-to-Tool. How agents access external capabilities.
- A2A (Agent-to-Agent): Agent-to-Agent. How agents negotiate and hand off work.
- OTP (Organization Transport Protocol): Org-to-Intelligence. How humans, agents, AI instances, and SOPs are wired together inside an organization, and shared across them.
MCP and A2A move bytes. OTP moves accountability. The three layers are complementary. Learn more about the stack.
Built by the System It Measures
OTP was built using the same AI agent coordination system it is designed to measure. The platform was constructed in 48 hours by a team of AI agents following the rules documented in their own OOS. This is recursive credibility: the system works because it was built by the system. Read the build story.
The Founder
David runs Sneeze It, a digital agency powered by 14 AI agents and a human team in production. He built OTP after discovering that the coordination between his humans and his agents was more valuable than any individual agent. His background spans digital advertising, call center operations, and AI systems architecture. He believes every organization will run on a hybrid human-and-AI operating system, and that the SOPs each human writes today are the inheritance every AI agent under them runs on tomorrow.
Map your team. Author the SOPs. Watch the AI agents under each human inherit them.