Aragog: Just-in-Time Model Routing for Scalable Serving of Agentic Workflows

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dai, Yinwei, Chen, Zhuofu, Iyer, Anand, Netravali, Ravi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912759363403776
author Dai, Yinwei
Chen, Zhuofu
Iyer, Anand
Netravali, Ravi
author_facet Dai, Yinwei
Chen, Zhuofu
Iyer, Anand
Netravali, Ravi
contents Agentic workflows have emerged as a powerful paradigm for solving complex, multi-stage tasks, but serving them at scale is computationally expensive given the many LLM inferences that each request must pass through. Configuration selection, or the cost-aware assignment of workflow agents to specific LLMs, can reduce these costs, but existing approaches bind configuration decisions before request execution, making them ill-suited for the heterogeneous and lengthy execution of workflows. Specifically, system loads can fluctuate rapidly and substantially during a request's lifetime, causing fixed configurations to quickly become suboptimal. We present Aragog, a system that progressively adapts a request's configuration throughout its execution to match runtime dynamics. To make this practical despite the massive space of workflow configurations, Aragog decouples the problem into two core elements -- a one-time routing step that identifies all accuracy-preserving configurations, and a cheap per-stage scheduler that selects among them using up-to-date system observations -- and introduces novel strategies to accelerate each. Across diverse workflows and model families, Aragog increases maximum serving throughput by 50.0--217.0\% and reduces median latency by 32.5--78.9\% at peak request rates, while maintaining accuracy comparable to the most expensive configurations.
format Preprint
id arxiv_https___arxiv_org_abs_2511_20975
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Aragog: Just-in-Time Model Routing for Scalable Serving of Agentic Workflows
Dai, Yinwei
Chen, Zhuofu
Iyer, Anand
Netravali, Ravi
Distributed, Parallel, and Cluster Computing
Agentic workflows have emerged as a powerful paradigm for solving complex, multi-stage tasks, but serving them at scale is computationally expensive given the many LLM inferences that each request must pass through. Configuration selection, or the cost-aware assignment of workflow agents to specific LLMs, can reduce these costs, but existing approaches bind configuration decisions before request execution, making them ill-suited for the heterogeneous and lengthy execution of workflows. Specifically, system loads can fluctuate rapidly and substantially during a request's lifetime, causing fixed configurations to quickly become suboptimal. We present Aragog, a system that progressively adapts a request's configuration throughout its execution to match runtime dynamics. To make this practical despite the massive space of workflow configurations, Aragog decouples the problem into two core elements -- a one-time routing step that identifies all accuracy-preserving configurations, and a cheap per-stage scheduler that selects among them using up-to-date system observations -- and introduces novel strategies to accelerate each. Across diverse workflows and model families, Aragog increases maximum serving throughput by 50.0--217.0\% and reduces median latency by 32.5--78.9\% at peak request rates, while maintaining accuracy comparable to the most expensive configurations.
title Aragog: Just-in-Time Model Routing for Scalable Serving of Agentic Workflows
topic Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2511.20975