SheetMind: An End-to-End LLM-Powered Multi-Agent Framework for Spreadsheet Automation

2025-06-14

AI Summary

Existing spreadsheet automation methods require users to possess formula or scripting knowledge, limiting accessibility for non-programmers. Method: We propose a modular multi-agent framework powered by large language models (LLMs), enabling end-to-end natural language–driven spreadsheet automation. Our novel three-agent architecture—comprising Manager, Executor, and Reflector agents—integrates BNF-grammar-constrained action generation with intent alignment verification to ensure interpretability, controllability, and robustness. The system is deeply integrated with Google Workspace as a production-ready extension. Contribution/Results: This work introduces the first grammar-guided multi-agent paradigm for spreadsheet automation, balancing practicality, reliability, and usability. On benchmark tasks, it achieves 80% success rate on single-step operations and 70% on multi-step workflows—substantially outperforming ablated variants and state-of-the-art baselines.

Technology Category

• Multiagent Systems: Teamwork
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Search and Retrieval-Augmented AI: Agentic search

Abstract

We present SheetMind, a modular multi-agent framework powered by large language models (LLMs) for spreadsheet automation via natural language instructions. The system comprises three specialized agents: a Manager Agent that decomposes complex user instructions into subtasks; an Action Agent that translates these into structured commands using a Backus Naur Form (BNF) grammar; and a Reflection Agent that validates alignment between generated actions and the user's original intent. Integrated into Google Sheets via a Workspace extension, SheetMind supports real-time interaction without requiring scripting or formula knowledge. Experiments on benchmark datasets demonstrate an 80 percent success rate on single step tasks and approximately 70 percent on multi step instructions, outperforming ablated and baseline variants. Our results highlight the effectiveness of multi agent decomposition and grammar based execution for bridging natural language and spreadsheet functionalities.

Problem

• Automates spreadsheet tasks via natural language instructions
• Decomposes complex instructions into executable subtasks
• Validates alignment between actions and user intent

Innovation

• LLM-powered multi-agent framework for automation
• BNF grammar translates instructions to commands
• Real-time interaction via Google Sheets extension