Investment Management
A synthetic investment committee matched the human team’s decisions in 85% of evaluated cases and explained its reasoning.
The investment division of a global conglomerate wanted to evaluate more opportunities without increasing headcount. Leadership also wanted to codify how its strongest investors assessed founders and businesses.
The Experiment
ArcticBlue developed a synthetic investment committee using pitch decks, founder interview transcripts, investment memoranda, and historical company outcomes. These materials provided the context for evaluating opportunities and articulating investment reasoning.
We tested the committee against real investment decisions in iterative batches. Comparing its judgments with the human team’s decisions made agreement measurable and allowed the team to examine the rationale behind each assessment.
Results
The synthetic committee matched the human investment team’s decisions in 85% of evaluated cases. It also provided clearer explanations of the factors behind a founder’s potential success or failure.
The benchmark made agreement with human decisions measurable and gave the team a concrete basis for refining the approach.
Why it mattered
The experiment made investment reasoning more explicit and gave leadership a practical way to explore how experienced judgment could support a larger volume of assessments. The rationale remained available for human review.
