ASGE-RR: Agentic Service Graph Embedding with Revisable Reservations for Dynamic AI-Agent Calls

2026-08-06

AI Summary

This work addresses runtime resource allocation conflicts in dynamic AI agent workflows, where exposed service dependencies can lead to premature resource consumption and failure to complete high-value tasks. The authors model agent invocations as a dynamic service graph embedding problem and propose a revisable reservation mechanism that integrates workflow continuation prediction to jointly map service replicas and network paths online while dynamically updating resource reservations. This approach proactively safeguards resources required for future high-value tasks under capacity, cost, and deadline constraints. WAN-based experiments demonstrate up to a 10% improvement in workflow completion value over baseline methods, validating the effectiveness of runtime structure-aware resource scheduling.

Abstract

AI-agent workflows often involve remote calls to models, memory stores, and tools distributed across a network. As execution progresses, these dependency calls collectively form an agentic service graph (ASG). Unlike traditional service requests, many dependency calls are revealed only at runtime. Consequently, allocating resources to a currently visible call may consume capacity later needed by a call from a higher-value workflow. We formulate this challenge as Agentic Service Graph Embedding (ASGE), an online network-control problem that maps runtime-revealed workflow calls to service replicas and network paths under capacity, cost and deadline constraints. We present ASGE-RR, an online ASGE controller with revisable reservations. ASGE-RR protects capacity for likely future calls while enforcing the constraints. ASGE-RR evaluates candidate replica-and-path mappings against predicted workflow continuations and updates reservations as new execution information becomes available. We evaluate ASGE-RR using OpenHands and GPT Researcher workflows executed with gpt-5.6-luna and replayed over in two complementary experimental environments, a controlled Docker testbed and a WAN testbed. The investigation shows that all the evaluated AI-agent tasks expose at least one runtime-revealed dependency call that can be steered before connection establishment. Exploiting this control point, even though the experimental environments are small-scale, ASGE-RR already demonstrates noticeable potential: It completes (up to) 10% more workflow value than a same-information rolling-horizon controller and a current-call steering controller on the WAN testbed. The results suggest that runtime-revealed workflow structure creates a new network control opportunity: protecting resources for likely future calls allows more AI-agent workflows to finish in time.

Problem

• Agentic Service Graph Embedding
• Dynamic AI-Agent Calls
• Runtime-Revealed Dependencies
• Resource Reservation
• Online Network Control

Innovation

• Agentic Service Graph
• Revisable Reservations
• Runtime-revealed Dependencies
• Online Network Control
• AI-Agent Workflows