We piloted agentic AI to replicate a moderately complex SLR, successfully identifying 14 of the 24 target studies. While the AI demonstrated significant potential for accelerating the search, a performance gap remains. This experiment highlights that expert oversight is still vital to refine results and fill critical search gaps. Ultimately, AI serves as a powerful partner, but human rigor ensures the integrity of the final review.
We deployed agentic AI to replicate a moderately complex economic model, automating data integration and variable calibration. The AI successfully streamlined the modeling workflow, yet required manual adjustments to capture subtle market nuances. This demonstrates that expert oversight is essential to validate assumptions and fill the gaps in automated reasoning. While AI accelerates the process, human expertise remains the anchor for economic accuracy and reliability