Hunt Globally: Deep Research AI Agents for Drug Asset Scouting in Investing, Business Development, and Search & Evaluation.

By: Alisa Vinogradova, Vlad Vinogradov, Luba Greenwood, Ilya Yasny, Dmitry Kobyzev, Shoman Kasbekar, Kong Nguyen, Dmitrii Radkevich, Roman Doronin, Andrey Doronichev

Published: 2026-02-16

View on arXiv →
#cs.AI

Abstract

Bio-pharmaceutical innovation has shifted, with most new drug assets originating outside the U.S. and disclosed through non-English channels. This creates multi-billion-dollar risks for investors and development teams who struggle to find "under-the-radar" assets using current Deep Research AI agents. This paper proposes a benchmarking methodology and a tuned, tree-based self-learning Bioptic Agent to achieve complete, non-hallucinated scouting across heterogeneous, multilingual sources. The Bioptic Agent significantly outperforms other models in F1 score.

FEEDBACK

Projects

No projects yet