Assembly

Assembly is a multi‑agent AI decision simulator that models how real people might react to a proposed action or product before it is launched. It ingests large volumes of publicly available text—news articles, forum posts, social‑media discussions—to create a pool of AI agents each embodying a distinct real‑world persona. Those agents debate a user‑provided question over several rounds, allowing opinions to shift, converge, or remain divided.

The system then generates a “God’s Eye View” report summarizing the distribution of positions (for, against, neutral), the arguments that proved most influential, and a confidence score for the predicted outcome. Two operational modes share the same simulation core: a Public Opinion Lab for policy, campus, or community decisions, and a Launch Lab aimed at direct‑to‑consumer brands seeking insight into pricing, product appeal, and competitive response.

Assembly is positioned for users who need data‑driven forecasts of public sentiment without conducting traditional surveys, offering a real‑time view of how opinions may evolve in response to a specific proposal.