TimeART: Towards Agentic Time Series Reasoning via Tool-Augmentation

2026-01-20

AI Summary

This work addresses the longstanding reliance on human experts in time series analysis, which has hindered automation and intelligent reasoning. To bridge this gap, we propose the first autonomous agent framework tailored for time series question answering, integrating the reasoning capabilities of large language models with specialized analytical tools. Built upon the TimeToolBench expert trajectory dataset, our approach employs a four-stage self-reflection training strategy to enable tool-augmented autonomous reasoning. Experimental results demonstrate that the proposed method achieves state-of-the-art performance across multiple time series question-answering benchmarks, thereby validating the feasibility and effectiveness of agent-based time series analysis.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Time series data widely exist in real-world cyber-physical systems. Though analyzing and interpreting them contributes to significant values, e.g, disaster prediction and financial risk control, current workflows mainly rely on human data scientists, which requires significant labor costs and lacks automation. To tackle this, we introduce TimeART, a framework fusing the analytical capability of strong out-of-the-box tools and the reasoning capability of Large Language Models (LLMs), which serves as a fully agentic data scientist for Time Series Question Answering (TSQA). To teach the LLM-based Time Series Reasoning Models (TSRMs) strategic tool-use, we also collect a 100k expert trajectory corpus called TimeToolBench. To enhance TSRMs'generalization capability, we then devise a four-stage training strategy, which boosts TSRMs through learning from their own early experiences and self-reflections. Experimentally, we train an 8B TSRM on TimeToolBench and equip it with the TimeART framework, and it achieves consistent state-of-the-art performance on multiple TSQA tasks, which pioneers a novel approach towards agentic time series reasoning.

Problem

• time series reasoning
• automation
• tool-augmentation
• Large Language Models
• TSQA

Innovation

• Agentic Reasoning
• Tool-Augmentation
• Time Series Question Answering
• Large Language Models
• Self-Reflection Training