Leveraging RAG for Enhanced Business Intelligence with Local LLMs
Bibliographic record
Abstract
This research addresses critical limitations in current business intelligence (BI) insights generation. It focuses on real-time data updates, accuracy, data privacy, security and more to build a system that is practically viable for modern-day businesses. Leveraging the proposed solution gives organizations a competitive edge in the fast-paced market. Research leverages emerging Large Language Models (LLMs) to derive relevant actionable insights while eliminating typical hallucinations. The research proposes system architecture that combines Retrieval-Augmented Generation (RAG) technology with local LLMs and real-time data streaming via Kafka to ensure data privacy, factual accuracy, and timely insights. Research presents quantitative and qualitative analysis for insights generation based on RAG vs. Direct LLM request. Moreover, research is executed on two different families of LLMs - thinking models and traditional models for detailed validation. Empirical testing reveals that RAG-based approaches outperform direct LLM queries in response time, CPU efficiency, and factual accuracy for insights generation. By leveraging local LLMs instead of cloud-based solutions, research aims to protect data and intellectual property of organizations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.013 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.008 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".