Reproduce “Agentic Retrieval of Topics and Insights from Earnings Calls” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Anant Gupta, Rajarshi Bhowmik, Geoffrey Gunow
Researchers at Bloomberg, USA, developed an LLM-agent driven approach to dynamically identify and structure emerging topics from corporate earnings call transcripts. This system constructs a hierarchical topic ontology that provides financial analysts with timely and actionable insights into evolving industry trends and company-specific narratives.
Key insights
- LLM-agentic systems can effectively overcome limitations of traditional topic modeling by dynamically identifying and structuring relevant, emerging topics within financial discourse.
- Organizing extracted topics into a hierarchical ontology significantly enhances knowledge management and allows for more nuanced tracking of evolving financial narratives.
- The system's ability to detect trends and emerging topics immediately after earnings calls provides timely insights, often appearing earlier than mainstream financial media.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.