Orchestrating Specialized Agents for Trustworthy Enterprise RAG

TL;DR

This paper introduces ADORE, a novel agentic framework for enterprise retrieval-augmented generation that improves trustworthiness, traceability, and completeness of generated reports through structured memory and iterative investigation.

Contribution

ADORE replaces linear retrieval with an iterative, user-steered investigation framework using a structured Memory Bank for improved traceability and evidence coverage in enterprise RAG tasks.

Findings

ADORE ranks first on DeepResearch Bench with a score of 52.65.

ADORE achieves a 77.2% head-to-head preference win rate on DeepConsult.

ADORE demonstrates superior evidence coverage and traceability in enterprise report generation.

Abstract

Retrieval-Augmented Generation (RAG) shows promise for enterprise knowledge work, yet it often underperforms in high-stakes decision settings that require deep synthesis, strict traceability, and recovery from underspecified prompts. One-pass retrieval-and-write pipelines frequently yield shallow summaries, inconsistent grounding, and weak mechanisms for completeness verification. We introduce ADORE (Adaptive Deep Orchestration for Research in Enterprise), an agentic framework that replaces linear retrieval with iterative, user-steered investigation coordinated by a central orchestrator and a set of specialized agents. ADORE's key insight is that a structured Memory Bank (a curated evidence store with explicit claim-evidence linkage and section-level admissible evidence) enables traceable report generation and systematic checks for evidence completeness. Our contributions are threefold: (1) Memory-locked synthesis - report generation is constrained to a structured Memory Bank (Claim-Evidence Graph) with section-level admissible evidence, enabling traceable claims and grounded citations; (2) Evidence-coverage-guided execution - a retrieval-reflection loop audits section-level evidence coverage to trigger targeted follow-up retrieval and terminates via an evidence-driven stopping criterion; (3) Section-packed long-context grounding - section-level packing, pruning, and citation-preserving compression make long-form synthesis feasible under context limits. Across our evaluation suite, ADORE ranks first on DeepResearch Bench (52.65) and achieves the highest head-to-head preference win rate on DeepConsult (77.2%) against commercial systems.