ResearchAI/TechIndependent Researcher, Agentic AI Project
Multi-Agent Research and Script Lab
Research, script, review, and human-approval prototype now being developed into a formal technical report.
The problem
Agentic research pipelines can generate large volumes of drafted material with no reliable review or approval step, making their output hard to trust or publish.
The approach
- 01Split the pipeline into research, script, and review agent roles
- 02Require an explicit human-approval gate before any output is treated as final
- 03Document the architecture and behavior for a technical write-up
- 04Develop the write-up into a formal technical report with faculty writing guidance
Tools & stack
Python multi-agent orchestrationHuman-in-the-loop approval workflowTechnical research writing
Results
- → Working prototype live at multi-agent-lab.com
- → Technical report/manuscript in development, July 2026 to present
- → Advised by Dr. Zhaohu (Jonathan) Fan, Georgia Tech Scheller College of Business — independent research with faculty guidance, not a Georgia Tech position, sponsored project, or publication
- → A potential INFORMS submission depends on quality, maturity, and fit — not secured
Deliverables & links
Want to work together?
Let's talk about what you're building.
Get in touch →