Fan Out or Relay? Ownership-Aware Routing for LLM Agents Querying Siloed Tables
New in LLM agents and tool use; no direction yet. LLM agents and tool use
Abstract (arXiv)
Inside organizations, a question's data often sit with several owners, who query their own tables but never hand them over. An LLM coordinator can fan out, sending the question to every owner at once, or relay, querying owners in turn and reusing earlier rows. We show that neither is safe in advance, because data ownership decides which one is right. Across thirteen settings on PeopleJoin and BIRD with three model families, fan-out is more accurate in twelve on questions one owner can answer, yet answers at most 4.3% of those joining two owners' tables. The flip follows ownership, not query shape: on BIRD two-table joins, fan-out answers 32-44% when one owner holds both and 1-4% otherwise. Relay wins by conditioning on returned rows, and hiding them from an otherwise identical planner cuts two-owner accuracy from 23.2% to 5.1%. Since ownership is public, we propose Ownership-Aware Routing (OAR), which drafts a schema-only SQL query, counts the owners whose tables it joins, and routes the question before any private row is read. In ten settings OAR is never significantly less accurate than the best fixed protocol, and beats every fixed protocol on PeopleJoin, by up to 9.3 points on held-out organizations.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 18 of its 32 references that are in the map. First posted on arXiv 13 Oct 2026.
Builds on
18 of 32 references.
References · 18
- 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
- 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
- 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,784
- 2023CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Societycited by 1,395
- 2018Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Taskcited by 1,139
- 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,075
- 2022Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questionscited by 923
- 2023Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLscited by 688
- 2024Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexitycited by 451
- 2024Scaling Large Language Model-based Multi-Agent Collaborationcited by 241
- 2023Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theorycited by 230
- 2023An LLM Compiler for Parallel Function Callingcited by 179
6 more
- 2023MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQLcited by 164
- 2026DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compressioncited by 76
- 2025Sample-Efficient Online Learning in LM Agents via Hindsight Trajectory Rewritingcited by 13
- 2025AgentAsk: Multi-Agent Systems Need to Askcited by 10
- 2025LM Agents for Coordinating Multi-User Information Gatheringcited by 5
- 2024Cooperative SQL Generation for Segmented Databases By Using Multi-functional LLM Agentscited by 2
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2025EllieSQL: Cost-Efficient Text-to-SQL with Complexity-Aware Routing
- 2025DiSRouter: Distributed Self-Routing for LLM Selections
- 2024BEAVER: An Enterprise Benchmark for Text-to-SQL
- 2026SemJoin: Semantic Join Optimization
- 2024Harnessing the Power of Multiple Minds: Lessons Learned from LLM Routing
- 2025THOR: Transformer Heuristics for On-Demand Retrieval
- 2025Toward Real-World Table Agents: Capabilities, Workflows, and Design Principles for LLM-based Table Intelligence
- 2025Permissioned LLMs: Enforcing Access Control in Large Language Models