paper-with-me

홈 › Papers

Scalable Inference Architectures for Compound AI Systems: A Production Deployment Study

2026-04-28 · Srikanta Prasad S, Utkarsh Arora arxiv

Modern enterprise AI applications increasingly rely on compound AI systems - architectures that compose multiple models, retrievers, and tools to accomplish complex tasks. Deploying such systems in production demands inference infrastructure that can efficiently serve concurrent, heterogeneous model invocations while maintaining cost-effectiveness and low latency. This paper presents a production deployment study of a modular, platform-agnostic inference architecture developed at Salesforce to support compound AI use cases including Agentforce (autonomous AI agents) and ApexGuru (AI-powered code analysis). The system integrates serverless execution, dynamic autoscaling, and MLOps pipelines to deliver consistent low-latency inference across multi-component agent workflows. We report production results demonstrating over 50% reduction in tail latency (P95), up to 3.9x throughput improvement, and 30 to 40% cost savings compared to prior static deployments. We further present a novel analysis of compound-system-specific challenges including multi-model fan-out overhead, cascading cold-start propagation, and heterogeneous scaling dynamics that emerge uniquely when serving agentic workloads. Through detailed case studies and operational lessons, we illustrate how the architecture enables compound AI systems to scale model invocations in parallel, handle bursty multi-agent workloads, and support rapid model iteration - capabilities essential for operationalizing agentic AI at enterprise scale.

📄 PDF Abstract BibTeX arXiv:2604.25724

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

ModServe: Scalable and Resource-Efficient Large Multimodal Model Serving

2025-02-02 · Haoran Qiu, Anish Biswas, Zihan Zhao, Jayashree Mohan 외

Large multimodal models (LMMs) demonstrate impressive capabilities in understanding images, videos, and audio beyond text. However, efficiently serving LMMs in production environments poses significant challenges due to …

DecoderGPUmodelScheduling

Genflow Ad Studio: A Compound AI Architecture for Brand-Aligned, Self-Correcting Video Generation

2026-05-16 · Debanshu Das, Lavi Nigam, Sunil Kumar Jang Bahadur, Gopala Dhar arxiv

Recent advancements in generative video models demonstrate high visual fidelity, yet their integration into enterprise environments is restricted by temporal inconsistencies and severe brand misalignment. Current monolit…

Video Generation

QUIVER: A Formal Framework for Quantifying Perturbation Propagation and Bifurcation in Compound AI Systems

2026-05-11 · Prashanti Nilayam, Sankalp Nayak arxiv

Compound AI systems that chain multiple LLM calls into directed computation graphs are now the dominant architecture for production AI. Although these architectures leverage heterogeneous nodes with mixed-mode outputs, n…

Podracer architectures for scalable Reinforcement Learning

2021-04-13 · Matteo Hessel, Manuel Kroiss, Aidan Clark, Iurii Kemaev 외

Supporting state-of-the-art AI research requires balancing rapid prototyping, ease of use, and quick iteration, with the ability to deploy experiments at a scale traditionally associated with production systems.Deep lear…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)

A Blueprint Architecture of Compound AI Systems for Enterprise

2024-06-02 · Eser Kandogan, Sajjadur Rahman, Nikita Bhutani, Dan Zhang 외

Large Language Models (LLMs) have showcased remarkable capabilities surpassing conventional NLP challenges, creating opportunities for use in production use cases. Towards this goal, there is a notable shift to building …