Parallel and Multi-Stage Knowledge Graph Retrieval for Behaviorally Aligned Financial Asset Recommendations

TL;DR

This paper presents RAG-FLARKO, a multi-stage, parallel knowledge graph retrieval method that improves personalized financial asset recommendations by grounding LLMs in relevant behavioral and market data, reducing context overhead and enhancing performance.

Contribution

It introduces RAG-FLARKO, a scalable, retrieval-augmented framework that effectively filters and constructs relevant knowledge graphs for improved financial recommendations.

Findings

Significantly improves recommendation quality on real-world data.

Enables smaller models to achieve high profitability and behavioral alignment.

Reduces context overhead and enhances relevance in LLM-based financial advice.

Abstract

Large language models (LLMs) show promise for personalized financial recommendations but are hampered by context limits, hallucinations, and a lack of behavioral grounding. Our prior work, FLARKO, embedded structured knowledge graphs (KGs) in LLM prompts to align advice with user behavior and market data. This paper introduces RAG-FLARKO, a retrieval-augmented extension to FLARKO, that overcomes scalability and relevance challenges using multi-stage and parallel KG retrieval processes. Our method first retrieves behaviorally relevant entities from a user's transaction KG and then uses this context to filter temporally consistent signals from a market KG, constructing a compact, grounded subgraph for the LLM. This pipeline reduces context overhead and sharpens the model's focus on relevant information. Empirical evaluation on a real-world financial transaction dataset demonstrates that RAG-FLARKO significantly enhances recommendation quality. Notably, our framework enables smaller, more efficient models to achieve high performance in both profitability and behavioral alignment, presenting a viable path for deploying grounded financial AI in resource-constrained environments.