KairosAgent: Agentic Time Series Forecasting with Fused Semantic Reasoning

2026-05-28

AI Summary

Existing cross-domain multimodal time series forecasting methods struggle to simultaneously achieve precise numerical modeling, cross-domain semantic reasoning, and effective multimodal fusion. This work proposes the first agent-based framework that integrates a large language model–driven reasoner with a time series foundation model–driven predictor. By dynamically invoking external tools, the framework enhances both numerical comprehension and semantic reasoning, seamlessly incorporating reasoning outcomes into the prediction pipeline. The approach innovatively introduces a prediction-aware reinforcement learning paradigm alongside a high-quality, multi-turn reasoning trajectory dataset to optimize agent decision-making. Evaluated under a zero-shot setting, the method significantly outperforms current state-of-the-art models, yielding notable improvements in both forecasting accuracy and interpretability.

Abstract

Cross-domain multimodal time series forecasting is a challenging task, requiring models to integrate precise numerical comprehension, cross-domain semantic understanding, and effective multimodal fusion. Existing approaches either build Time Series Foundation Models (TSFMs) from scratch or leverage pretrained Large Language Models (LLMs). However, TSFMs often overlook semantic understanding and lack the ability to perform future-oriented semantic reasoning, and LLMs struggle with numerical comprehension and accurate quantitative forecasting. To overcome these limitations, we propose KairosAgent, a novel agentic framework for multimodal time series forecasting, including an LLM-based reasoner and a TSFM-based forecaster. KairosAgent unifies textual reasoning and numerical forecasting by dynamically invoking analytical tools to enhance the numerical understanding and semantic reasoning capabilities of LLMs. The reasoning results are subsequently fused into the TSFM pipeline, enabling more accurate and reliable future predictions. To further improve the reasoning, we curate a large-scale corpus of high-quality trajectories, alongside a reinforcement learning from forecasting paradigm with multi-turn refinement and turn-level credit assignment. Experiments demonstrate that KairosAgent achieves superior zero-shot forecasting performance while maximizing the utility of pretrained LLMs and TSFMs, presenting a promising direction for efficient and interpretable time series agents. The project page is at https://foundation-model-research.github.io/KairosAgent .

Problem

• time series forecasting
• multimodal fusion
• semantic reasoning
• numerical comprehension
• cross-domain

Innovation

• Agentic forecasting
• Multimodal time series
• Semantic-numerical fusion
• Reinforcement learning from forecasting
• Time Series Foundation Models