RAGSmith: A Framework for Finding the Optimal Composition of Retrieval-Augmented Generation Methods Across Datasets

TL;DR

RAGSmith is a modular framework that uses evolutionary search to optimize retrieval-augmented generation pipelines across diverse datasets, significantly improving performance over naive configurations.

Contribution

It introduces a comprehensive end-to-end architecture search method for RAG systems, optimizing over multiple modules and configurations with a genetic algorithm.

Findings

RAGSmith outperforms naive RAG baselines by +3.8% on average across domains.

The search explores about 0.2% of the configuration space, roughly 100 candidates.

A robust backbone of vector retrieval plus post-generation reflection is commonly selected.

Abstract

Retrieval-Augmented Generation (RAG) quality depends on many interacting choices across retrieval, ranking, augmentation, prompting, and generation, so optimizing modules in isolation is brittle. We introduce RAGSmith, a modular framework that treats RAG design as an end-to-end architecture search over nine technique families and 46{,}080 feasible pipeline configurations. A genetic search optimizes a scalar objective that jointly aggregates retrieval metrics (recall@k, mAP, nDCG, MRR) and generation metrics (LLM-Judge and semantic similarity). We evaluate on six Wikipedia-derived domains (Mathematics, Law, Finance, Medicine, Defense Industry, Computer Science), each with 100 questions spanning factual, interpretation, and long-answer types. RAGSmith finds configurations that consistently outperform naive RAG baseline by +3.8\% on average (range +1.2\% to +6.9\% across domains), with gains up to +12.5\% in retrieval and +7.5\% in generation. The search typically explores $\approx 0.2\%$ of the space ($\sim 100$ candidates) and discovers a robust backbone -- vector retrieval plus post-generation reflection/revision -- augmented by domain-dependent choices in expansion, reranking, augmentation, and prompt reordering; passage compression is never selected. Improvement magnitude correlates with question type, with larger gains on factual/long-answer mixes than interpretation-heavy sets. These results provide practical, domain-aware guidance for assembling effective RAG systems and demonstrate the utility of evolutionary search for full-pipeline optimization.