FinBloom: Knowledge Grounding Large Language Model with Real-time Financial Data

Paper: arXiv 2502.18471

Abstract

Large language models (LLMs) excel at generating human-like responses but often struggle with interactive tasks that require access to real-time information. This limitation poses challenges in finance, where models must access up-to-date information, such as recent news or price movements, to support decision-making. To address this, we introduce Financial Agent, a knowledge-grounding approach for LLMs to handle financial queries using real-time text and tabular data. Our contributions are threefold: First, we develop a Financial Context Dataset of over 50,000 financial queries paired with the required context. Second, we train FinBloom 7B, a custom 7 billion parameter LLM, on 14 million financial news articles from Reuters and Deutsche Presse-Agentur, alongside 12 million Securities and Exchange Commission (SEC) filings. Third, we fine-tune FinBloom 7B using the Financial Context Dataset to serve as a Financial Agent. This agent generates relevant financial context, enabling efficient real-time data retrieval to answer user queries. By reducing latency and eliminating the need for users to manually provide accurate data, our approach significantly enhances the capability of LLMs to handle dynamic financial tasks. Our proposed approach makes real-time financial decisions, algorithmic trading and other related tasks streamlined, and is valuable in contexts with high-velocity data flows.

Complexity vs Empirical Score

• Math Complexity: 2.0/10
• Empirical Rigor: 8.0/10
• Quadrant: Street Traders — practical and empirical, lighter on theory

Why this score: The paper focuses on implementing a practical LLM agent system using large-scale financial datasets and real-time data integration, with minimal advanced mathematical derivations.

Research Flowchart

  flowchart TD
  A["Research Goal: Enable LLMs to answer financial queries<br>using real-time data"] --> B{"Methodology"};
  B --> C["Develop Financial Context Dataset<br>50k queries"];
  B --> D["Train FinBloom 7B<br>on 14M news + 12M SEC filings"];
  B --> E["Fine-tune as Financial Agent<br>to generate retrieval contexts"];
  C --> F["Computational Process:<br>Agent retrieves real-time text & tabular data"];
  D --> F;
  E --> F;
  F --> G["Key Findings: Efficient real-time data retrieval<br>Reduced latency for financial decision-making<br>Streamlined algorithmic trading workflows"];

Cited References (66)

1. Language models are few-shot learners (2020)
2. Bloom: A 176b-parameter open-access multilingual language model (2022) — link
3. Opt: O pen pre-trained transformer language models (2022) — link
4. BloombergGPT : A Large Language Model for Finance (2023) — link
5. BioGPT : G enerative pre-trained transformer for biomedical text generation and mining (2022)
6. RAG Does Not Work for Enterprises (2024) — link
7. A Comprehensive Survey of Retrieval-Augmented Generation ( RAG ): E volution, Current Landscape and Future Directions (2024) — link
8. The power of noise: R edefining retrieval for RAG systems (2024)

2 of 66 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.