Researching how AI completes the work
WaoX Research studies two things: what AI can do now, and what AI can do next. The first becomes products; the second becomes direction.
What we research
From a single agent to an organization, four themes make up the Research Institute's current map.
AI runtime
Agent Runtime
The runtime where AI does the work: understanding goals, breaking down tasks, using tools, executing continuously and delivering results. It is the technical foundation of products like WaoxAI and AI Assistant, already running in production.
AI Organization OS
AI Organization OS
If a company is an operating system, AI Organization OS is its kernel: how roles, business processes, memory and decisions run with people and AI together.
OPC one-person company
One Person Company
Can one person plus N AI digital employees found and run a company? We are studying the smallest viable form of a one-person company.
Multi-agent collaboration
Multi-Agent
A single agent has limits. Multiple agents divide work, coordinate and check each other to accomplish what one AI cannot.
Current research results
The Research Institute's output has three layers: shipped products, accumulated capabilities and public technical foundations.
WaoxAI, AI Assistant, Creator Center, AIPPT, AI Meetings and GEO — the AI matrix and AI products are live and operating, the most direct expression of our research.
View productsAI Assistant embodies our research on memory and multi-step task execution: remembering preferences, connecting tasks and completing work in multiple steps.
Meet Your AI Work PartnerThe technical foundation powering our products: task planning, tool use (MCP), memory and retrieval-augmented generation (RAG).
View the technical architectureWhat is AI's runtime
A runtime is the foundation that makes a program run. An AI runtime is the mechanism that turns "understanding the goal" into "getting the work done".
Long-term plus short-term memory that retains preferences and context
Unified access to data, tools, and external services
Retrieval-augmented Q&A and creation grounded in your knowledge base
In one sentence: you give the AI a goal, and the runtime understands it, breaks it into steps, uses tools to execute and delivers the result to you.
What we're here to do: move toward AGI
AGI (artificial general intelligence) is our long-term goal, broken into three steps.
Now
AI gets work done
Single tasks and multi-step work: the AI matrix and AI products already "get the work done for you."
Recent
AI runs the organization
AI Organization OS, OPC one-person companies and multi-agent collaboration: from finishing a task to running an organization.
Future
AGI, artificial general intelligence
Giving AI general cross-domain capability is the Research Institute's long-term direction, and we are on our way.
From completing a task to running an organization to general capability — at every step, we build something usable first.
Researching how AI works —the journey continues
Whether you want to join the research team, share a research lead, or turn your business scenario into our next experiment, we would love to hear from you.