Tutorial video for Research Architect, an AI copilot for human-AI teaming research

Last month, we released Research Architect, an AI copilot for human-AI research collaboration focused on hypothesis framing. It is our first product in the human-AI teaming age, and it provides a shared workspace where human expertise, scientific literature, and AI reasoning come together to develop research plans.

There is no doubt that AI and AI agents have been deeply involved in scientific research. The cross-domain knowledge and outstanding coding skills of the frontier AI models are astonishing. However, before conducting solid research, framing a reasonable hypothesis for the research question requires more context and curated data. Research Architect gives users an interface to combine their contextual knowledge and expertise with AI’s broad knowledge and skills.

Researchers can explore relationships, examine their supporting explanations, define the study’s scope, and check data coverage. They can also give the AI explicit guidance, preserving important mechanisms even when current data cannot measure them.

One of our team members, Prof. Xiao Huang at Emory University, created a nice video tutorial on how to use Research Architect. You are welcome to watch this 3-minute video and try it on researcharchitect.ai for free now.

The video walks through a mosquito ecology example. Knowing that a container holds water is different from knowing how long the water remains. That distinction matters when studying mosquito development. Research Architect makes the missing measurement visible, connects literature to the relevant factor, and helps turn a candidate explanation into a study design involving repeated observations.

Research Architect is our first step in re-engineering scientific inquiry infrastructure for the AI era. We will continue developing Fractal Research Graph and Computational Worldspace to connect research knowledge and support shared modeling:

One Interface, Anyone. One Graph, Anything. One Model, Anywhere.

Watch the introduction and try bringing your own research question: researcharchitect.ai

Research Architect tutorial (3 minutes). Click to watch on LinkedIn.

More about this project: Research Architect.

First published on LinkedIn on September 14, 2026.