PLAIground: SLO-Driven Runtime Model Selection for Compound AI Systems in the Edge-Cloud-Space Continuum

2026-06-12

AI Summary

This work addresses the challenge of dynamically selecting models in composite AI systems across the edge-cloud-space continuum to simultaneously satisfy stringent service-level objectives (SLOs) concerning accuracy, latency, and cost. The authors propose Composable AI Models (CAIM), an abstraction that decouples task semantics from model implementation, and introduce Pixie, the first SLO-aware runtime model selection mechanism tailored for composite AI systems. Pixie dynamically schedules optimal models based on task and data contracts without requiring workflow modifications. Experimental evaluation demonstrates that Pixie achieves 91.3% accuracy on real-world workflows while meeting SLOs, reducing latency or cost violations by up to 21× compared to fixed-model strategies and incurring only 4% less accuracy loss.

Abstract

Applications in the 3D Computing Continuum, which unifies edge, cloud, and space, require combining multiple AI tasks such as object detection, time-series analytics, and natural language processing into Compound AI systems. These systems must satisfy stringent Service Level Objectives (SLOs) on accuracy, latency, and cost. A key mechanism for maintaining SLO compliance of Compound AI systems is runtime model selection, where AI models are dynamically switched for each workflow task. However, existing distributed and compound AI frameworks do not natively support runtime model selection. We present PLAIground, a framework that enables runtime model selection for Compound AI systems. PLAIground introduces Compoundable AI Model (CAIM) abstraction, which decouples task semantics from AI model implementations via Task and Data Contracts, enabling model switching without workflow changes. Additionally, PLAIground introduces Pixie, an SLO-driven runtime model selection algorithm, which dynamically selects the most suitable model for each task during execution. Our evaluation on two realistic Compound AI workflows demonstrates that Pixie achieves up to 91.3% accuracy while maintaining SLO compliance where fixed-model strategies either violate cost and latency budgets up to 21x or miss accuracy targets by 4%.

Problem

• Compound AI Systems
• Service Level Objectives
• Runtime Model Selection
• Edge-Cloud-Space Continuum
• Model Switching

Innovation

• runtime model selection
• Compound AI systems
• SLO-driven optimization
• CAIM abstraction
• edge-cloud-space continuum