# Rethinking AI Cloud Infrastructure for Agentic Serving Systems with the Aries Experimentation Framework

arXiv: https://arxiv.org/abs/2607.29069 · HTML: https://mlsanity.com/paper/2607.29069/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 31 Jul 2026.

## Built on by (most cited first, 2 of 2)

- [Taking Tokenization off the Host for Agentic LLM Serving](https://mlsanity.com/paper/2610.13325/) (2026)
- [Beyond Token Savings: A Systematic Study of Context Compression in LLM Agents](https://mlsanity.com/paper/2609.32961/) (2026)

## Abstract (arXiv)

Autonomous agents challenge conventional LLM serving by coupling repeated inference with persistent context and sandboxed tool execution. We present Aries, a full-stack experimentation framework that separates task semantics from execution configurations, reconstructs cross-component agent trajectories with correlated system telemetry, and exposes stateful tool execution through a consistent interface across heterogeneous sandbox substrates. We use Aries to conduct reproducible experiments on open agent harnesses and benchmarks. We complement these experiments with production traces from a commercial platform, grounding low-level systems research in observed production behavior. Our results show that (1) token-centric metrics miss non-inference bottlenecks, (2) retaining additional context yields diminishing accuracy benefits while reducing serving capacity, and (3) tool sandboxes alternate between long idle periods and short resource bursts, while current snapshot-based state management makes aggressive suspension costly. A complementary security analysis further highlights the need to reduce the sandbox attack surface. We then discuss the vision for agent-native serving systems designed around trajectory-level metrics, adaptive context management, elastic sandbox resource management, and sandboxes with minimized attack surface.